A method and system for broadband radar detection
Through Fourier transform and coherent accumulation technology, combined with Monte Carlo method, the problem of detection difficulty of broadband radar at low signal-to-noise ratio is solved, and high-accuracy target detection is achieved, and close to narrowband detection performance is achieved.
Patent Information
- Application Number
- CN202111594226.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-23
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2041-12-23
AI Technical Summary
Broadband radars are difficult to effectively detect targets under low signal-to-noise ratio, resulting in false detection and missed detection. Traditional narrowband point target detection is no longer applicable.
By setting the radar to transmit a linear frequency modulation signal, receiving and de-aberging echo processing, Fourier transform is performed, motion compensation and coherent accumulation are performed, and the Monte Carlo method and energy accumulation technology are used to reduce the influence of signal-to-noise ratio and improve detection accuracy.
In the case of low signal-to-noise ratio, the target detection performance is achieved close to narrowband detection performance, which improves detection accuracy and signal-to-noise ratio, and reduces false detection and missed detection.
Smart Images

Figure CN114265052B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of radar detection technology, and in particular to a method and system for broadband radar detection. Background Art
[0002] Wideband radar is a radar system with a relatively wide operating bandwidth and can be used in the field of target detection, that is, detecting whether a target exists.
[0003] Modern high-resolution radars can obtain higher resolution and more target information by transmitting broadband waveforms. When the target size is larger than the radar's resolution unit, the scattering points will be distributed on each range unit, and the target will be regarded as a range-extended target. Due to the dispersion of echo energy, the signal-to-noise ratio of a single range unit is reduced, and traditional narrowband point target detection is no longer applicable in radar systems.
[0004] The target echo of broadband radar is a synthesis of multiple scattering point echoes. The amplitude, frequency and phase of each scattering point echo are random, and the target information cannot be directly obtained on the range image, which affects the detection performance of broadband radar. When the signal-to-noise ratio is high and the target moves little along the radar line of sight, the energy of the echo spectrum can be directly accumulated. Although it will cause a certain loss of signal-to-noise ratio, it can still effectively detect the target; when the signal-to-noise ratio is low and the target moves faster, the distance unit where the target scattering point is located will be offset, and it is impossible to achieve effective detection effect directly through the energy accumulation of the echo, which is prone to false detection and missed detection. Summary of the invention
[0005] In order to solve the problems existing in the prior art, the object of the present invention is to provide a detection method and system for a wideband radar, which can achieve target detection under a lower signal-to-noise ratio, and the detection performance is close to that of narrowband detection.
[0006] To achieve the above object, the technical solution adopted by the present invention is: a method for broadband radar detection, comprising the following steps:
[0007] S1. Assume that the radar transmits a linear frequency modulation signal s(t). After the transmitted signal is reflected by the target, when the motion parameters of the target are constant within the pulse width, the echo received by the antenna end is recorded as r(t) after de-skewing.
[0008] S2. Perform Fourier transform on the broadband radar echo signal r(t) to obtain a high-resolution one-dimensional range image R(f) of the echo;
[0009] S3, performing motion compensation on the high-resolution one-dimensional range image of the received echo according to the frequency offset caused by the target motion, so that the range units corresponding to the strong scattering points in the range images of different echoes are aligned;
[0010] S4. Perform coherent accumulation on the M aligned adjacent range images, and intuitively estimate the number of strong scattering points through the peak number and position of the range image after coherent accumulation. and its corresponding distance unit;
[0011] S5, accumulating energy of adjacent high-resolution one-dimensional range images to obtain an energy focusing signal;
[0012] S6, using the Monte Carlo method to perform detection, using noise as the reference signal to substitute into the accumulation process of step S5, obtaining the noise energy focused signal and the maximum value of the noise energy focused signal, and according to the given false alarm probability P FA Repeat this step 1 / P FA times, we will get 1 / P FA The maximum value of the noise energy focusing signal is 1 / P FA The maximum values are arranged in descending order, and the value between the first value and the second value in the sorting result is taken as the detection threshold; then the maximum value of the energy focusing signal obtained in step S5 is compared with the detection threshold, if the former is greater than the latter, it is judged that there is a target, otherwise it is not.
[0013] The present invention also provides a system for broadband radar detection, comprising:
[0014] The receiving and processing module at the antenna end assumes that the radar transmits a linear frequency modulation signal s(t). After the transmitted signal is reflected by the target, when the motion parameter of the target is constant within the pulse width, the receiving and processing module at the antenna end de-skews the received echo and records it as r(t);
[0015] Fourier transform module, used to perform Fourier transform on the broadband radar echo signal r(t) to obtain a high-resolution one-dimensional range image R(f) of the echo;
[0016] A compensation module is used to perform motion compensation on the high-resolution one-dimensional range image of the received echo according to the frequency offset caused by the target motion, so that the range units corresponding to the strong scattering points in the range images of different echoes are aligned;
[0017] The coherent accumulation module is used to coherently accumulate M aligned adjacent range images, and intuitively estimate the number of strong scattering points through the peak number and position of the range image after coherent accumulation. and its corresponding distance unit;
[0018] An energy accumulation module, used to accumulate energy of adjacent high-resolution one-dimensional range images to obtain an energy focusing signal;
[0019] The detection module uses the Monte Carlo method for detection, uses noise as the reference signal to substitute into the accumulation process of the energy accumulation module, obtains the noise energy focusing signal and the maximum value of the noise energy focusing signal, and according to the given false alarm probability P FA Repeat this step 1 / P FA times, we will get 1 / P FA The maximum value of the noise energy focusing signal is 1 / P FA The maximum values are arranged in descending order, and the value between the first value and the second value in the sorting result is taken as the detection threshold; then the maximum value of the energy focusing signal obtained by the energy accumulation module is compared with the detection threshold, if the former is greater than the latter, it is judged that there is a target, otherwise it is not.
[0020] Compared with the prior art, the beneficial effects achieved by the present invention include: the present invention estimates the position of the strong scattering point by coherent accumulation of the range image, and then performs energy focusing on the strong scattering point of the high-resolution one-dimensional range image of the aligned adjacent echoes, thereby greatly improving the detection capability of the target. At the same time, the present invention makes full use of the information of the scattering point, and only performs conjugate multiplication on the distance unit corresponding to the strong scattering point in the detection window containing the strong scattering point, which effectively reduces the influence of the signal-to-noise ratio on the target detection, thereby improving the detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 is a flow chart of broadband radar detection in an embodiment of the present invention;
[0022] Figure 2 It is a schematic diagram comparing the performance of broadband detection and narrowband detection in an embodiment of the present invention. DETAILED DESCRIPTION
[0023] The technical solution of the present invention is further described in detail below in conjunction with embodiments and drawings, but the implementation manner of the present invention is not limited thereto.
[0024] Example 1
[0025] like Figure 1 As shown, the broadband radar detection method of this embodiment specifically includes the following steps:
[0026] Step 1: Assume that the radar transmits a linear frequency modulation signal s(t). After the transmitted signal is reflected by the target, when the motion parameters of the target are constant within the pulse width, the echo received by the antenna end is recorded as r(t) after de-skewing.
[0027] Step 2: Perform Fourier transform on the broadband radar echo signal r(t) to obtain a high-resolution one-dimensional range image R(f) of the echo.
[0028] Step 3: Perform motion compensation on the high-resolution one-dimensional range image of the received echo according to the frequency offset caused by the target motion, so that the range units corresponding to the strong scattering points in different echo range images can be basically aligned.
[0029] Step 4: Coherently accumulate the M aligned adjacent range images. The power of the range image does not change before and after the coherent accumulation, while the power of the noise decreases, which improves the signal-to-noise ratio by 10lgM dB. The number and position of the peaks of the range image after coherent accumulation can be used to intuitively estimate the number of strong scattering points. and its corresponding distance unit.
[0030] Step 5: Accumulate energy of adjacent high-resolution one-dimensional range images to obtain energy focusing signals. Step 5 specifically includes the following steps:
[0031] Step 5.1: Let the i-th high-resolution one-dimensional range image be denoted as R i (f), the i+1th high-resolution one-dimensional range image is denoted as R i+1 (f) According to the length of the target to be detected, a detection window with a length of L is selected, that is, the length of the detection window should be greater than the number of distance units occupied by the target, so that the strong scattering points can be completely distributed in one detection window;
[0032] Step 5.2: Sliding windows and accumulating the two adjacent one-dimensional range images of step 5.1. The sliding window means that the leading edge position of the detection window is moved to the trailing edge position of the detection window during the previous accumulation during the next accumulation. The accumulation means operating the scattering points in the detection window. If the strong scattering points are distributed in different detection windows, it is necessary to perform frequency shift on all the range images to be detected so that all the strong scattering points are distributed in the same detection window before sliding windows and accumulating.
[0033] The specific accumulation process is: if there is no strong scattering point distribution in the detection window, then all distance units of the i-th and i+1-th distance images in the detection window are conjugate multiplied and averaged, that is, When the detection window moves to the When the strong scattering point is located, only the strong scattering points of the i-th and i+1-th range images in the detection window are conjugate multiplied and averaged, that is, In this way, the energy of the scattering points can be accumulated and the influence of noise on detection can be reduced.
[0034] If it is difficult to obtain the distance unit of the strong scattering point, you can take a few more distance units when focusing the energy during the accumulation process in step 5.2. This will not have a significant impact on the detection performance.
[0035] Step 6: Use the Monte Carlo method for detection: Use noise as the reference signal and substitute it into the accumulation process of step 5.2 to obtain the noise energy focusing signal and the maximum value of the noise energy focusing signal. According to the given false alarm probability P FA Repeat this step (1 / P FA ) times, we will get (1 / P FA ) noise energy focusing signal maximum value, for (1 / P FA ) maximum values are arranged in descending order, and the value between the first value and the second value in the sorting result is taken as the detection threshold; then the maximum value of the energy focusing signal obtained in step 5 is compared with the detection threshold, if the former is greater than the latter, it is judged that there is a target, otherwise it is not.
[0036] Step 7. Since the corresponding distance unit remains unchanged after the range image is aligned, when the antenna receives the next echo, it only needs to repeat the accumulation process of step 5.2 and the detection process of step 6 with the range image of the next echo and the previous range image to detect whether there is a target.
[0037] To fully describe the technical solution of the present invention, an embodiment is listed below for illustration.
[0038] Assume that there is a target consisting of 17 ideal points, the radial velocity of the target to the radar is 300 m / s, the center frequency of the radar is 9 GHz, the sampling frequency is 1 GHz, the pulse width is 100 μs, the bandwidth of the wideband waveform is set to 1 GHz, and the bandwidth of the narrowband waveform is set to 5 MHz. The size of the generated de-skewed wideband echo is 12000, and the size of the narrowband echo is 600.
[0039] If the signal-to-noise ratio during CFAR detection for narrowband signals is SNR N =SNR, that is, the noise variance that needs to be added in narrowband detection is Where E N is the narrowband signal energy, then the signal-to-noise ratio of the signal to noise on the broadband time domain echo should be SNR W =SNR-10lg12000. Based on the signal-to-noise ratio, the noise variance of the broadband echo can be calculated as Where E w is the narrowband signal energy.
[0040] For several broadband radar echoes containing noise, motion compensation alignment is performed according to the frequency offset caused by target motion. The range image is coherently accumulated by Fourier transform of the echo to obtain 7 strong scattering points. The estimated relative positions of the strong scattering points are 366, 399, 410, 418, 426, 439, and 470.
[0041] According to the estimated relative position of the strong scattering point, the size of the detection window is L = 120, and starting from the first distance unit, the detection window 120 is used as the step length, and the 120 distance units of R0(f) and R1(f) in the detection window are conjugate multiplied according to step 5.2. When the detection window slides to (361, 480), only the distance units corresponding to the above 7 scattering points of R0(f) and R1(f) in the detection window are conjugate multiplied and added to obtain the average, so that the energy focusing signal can be obtained.
[0042] Set the false alarm probability to 10 -3 , only for noise variance σ 2 W Repeat step 5.2 for 1000 times to get the detection threshold. Compare the maximum value of the energy focusing signal detected each time with the detection threshold to determine whether the target exists. For narrowband echoes, directly use matched filtering on narrowband echoes with noise, and then perform CFAR detection to determine whether the target exists. Figure 2 The detection performance of wideband and narrowband, the result of wideband detection is slightly lower than that of narrowband but the difference is not big.
[0043] It can be seen that in this embodiment, a high-resolution range image is obtained after Fourier transform of the received radar echo signal, and a number of adjacent range images after motion compensation are coherently accumulated to determine the number of strong scattering points and the range units in which they are located. A detection window of appropriate size is determined according to the length of the target to be detected, and then sliding window energy accumulation is performed on the two aligned adjacent range images. In the detection window containing strong scattering points, only the strong scattering points are conjugate multiplied and accumulated, and in the detection window without strong scattering points, all the range units therein are conjugate multiplied and accumulated, and the maximum value of the accumulated focused signal is compared with the detection threshold. Since this embodiment makes full use of the position information of the strong scattering points, it effectively reduces the influence of the signal-to-noise ratio on target detection, and its detection performance is close to that of narrowband detection.
[0044] Example 2
[0045] Based on the same inventive concept as Example 1, this embodiment is a system for broadband radar detection, which specifically includes the following modules:
[0046] The antenna end receiving and processing module is used to implement step 1 of embodiment 1. Assume that the radar transmits a linear frequency modulation signal s(t). After the transmitted signal is reflected by the target, when the motion parameter of the target is constant within the pulse width, the antenna end receiving and processing module de-skews the echo received by the signal and records it as r(t).
[0047] A Fourier transform module, used to implement step 2 of embodiment 1, to perform Fourier transform on the broadband radar echo signal r(t) to obtain a high-resolution one-dimensional range image R(f) of the echo;
[0048] A compensation module, used to implement step 3 of embodiment 1, to perform motion compensation on the high-resolution one-dimensional range image of the received echo according to the frequency offset caused by the target motion, so that the range units corresponding to the strong scattering points in the range images of different echoes are aligned;
[0049] A coherent accumulation module is used to implement step 4 of embodiment 1, to coherently accumulate M aligned adjacent range images, and to intuitively estimate the number of strong scattering points through the peak number and position of the range image after coherent accumulation. and its corresponding distance unit;
[0050] An energy accumulation module, used to implement step 5 of embodiment 1, to accumulate energy of adjacent high-resolution one-dimensional range images to obtain an energy focusing signal;
[0051] The detection module implements step 6 of embodiment 1, uses the Monte Carlo method to perform detection, uses noise as a reference signal to substitute into the accumulation process of the energy accumulation module, obtains the noise energy focusing signal and the maximum value of the noise energy focusing signal, and according to the given false alarm probability P FA Repeat this step 1 / P FA times, we will get 1 / P FA The maximum value of the noise energy focusing signal is 1 / P FA The maximum values are arranged in descending order, and the value between the first value and the second value in the sorting result is taken as the detection threshold; then the maximum value of the energy focusing signal obtained by the energy accumulation module is compared with the detection threshold, if the former is greater than the latter, it is judged that there is a target, otherwise it is not.
[0052] In addition, when the antenna end receiving processing module receives the next echo, the range image of the next echo and the previous range image repeat the accumulation process of the energy accumulation module and the detection process of the detection module to detect whether there is a target.
[0053] The above description is merely an embodiment of the present invention and does not limit the technical scope of the present invention. Therefore, any minor modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention are still within the scope of the technical solution of the present invention.
Claims
1. A method for broadband radar detection, characterized in that: The following steps are involved: S1. Assume that the radar transmits a linear frequency modulation signal s(t). After the signal is reflected by the target, when the target's motion parameters are constant within the pulse width, the echo received by the antenna end is recorded as r(t) after de-skewing. S2. Perform Fourier transform on the broadband radar echo signal r(t) to obtain a high-resolution one-dimensional range image R(f) of the echo; S3, performing motion compensation on the high-resolution one-dimensional range image of the received echo according to the frequency offset caused by the target motion, so that the range units corresponding to the strong scattering points in the range images of different echoes are aligned; S4. Perform coherent accumulation on the M aligned adjacent range images, and intuitively estimate the number of strong scattering points through the peak number and position of the range image after coherent accumulation. and its corresponding distance unit; S5, accumulating energy of adjacent high-resolution one-dimensional range images to obtain an energy focusing signal; S6, using the Monte Carlo method to perform detection, using noise as the reference signal to substitute into the accumulation process of step S5, obtaining the noise energy focused signal and the maximum value of the noise energy focused signal, and according to the given false alarm probability P FA Repeat this step 1 / P FA times, we will get 1 / P FA The maximum value of the noise energy focusing signal is 1 / P FA The maximum values are arranged in descending order, and the value between the first value and the second value in the sorting result is taken as the detection threshold; then the maximum value of the energy focusing signal obtained in step S5 is compared with the detection threshold, and if the former is greater than the latter, it is judged that there is a target, otherwise it is not; Step S5 includes: S51, let the i-th high-resolution one-dimensional range image be denoted by R i (f), the i+1th high-resolution one-dimensional range image is denoted as R i+1 (f); According to the length of the target to be detected, a detection window with a length of L is selected, and the length of the detection window is greater than the number of distance units occupied by the target, so that the strong scattering points are completely distributed in one detection window; S52, sliding windows and accumulating two adjacent one-dimensional range images of step S51; the sliding window means that the leading edge position of the detection window in the next accumulation is moved to the trailing edge position of the detection window in the previous accumulation, and the accumulation means operating the scattering points in the detection window. If the strong scattering points are distributed in different detection windows, frequency shifting is performed on all the range images to be detected so that all the strong scattering points are distributed in the same detection window, and then sliding windows and accumulating are performed; The accumulation process of step S52 is: If there is no strong scattering point distribution in the detection window, then all the distance units of the i-th and i+1-th range images in the detection window are conjugate multiplied and averaged. When the detection window moves to the When the position of a strong scattering point is determined, only the strong scattering points of the i-th and i+1-th range images in the detection window are conjugate multiplied and averaged. This allows the energy at the scattering points to be accumulated.
2. The method for broadband radar detection according to claim 1, characterized in that: The method further comprises the steps of: S7. When the antenna receives the next echo, the accumulation process of step S5 and the detection process of step S6 are repeated for the range image of the next echo and the previous range image to detect whether there is a target.
3. The method for broadband radar detection according to claim 1, characterized in that: The size of the detection window is L=120.
4. A broadband radar detection system, characterized in that: include: Antenna receiving and processing module, assuming that the radar transmits a linear frequency modulation signal s(t). After the signal is reflected by the target, when the target's motion parameters are constant within the pulse width, the antenna receiving and processing module de-skews the received echo and records it as r(t); Fourier transform module, used to perform Fourier transform on the broadband radar echo signal r(t) to obtain a high-resolution one-dimensional range image R(f) of the echo; A compensation module is used to perform motion compensation on the high-resolution one-dimensional range image of the received echo according to the frequency offset caused by the target motion, so that the range units corresponding to the strong scattering points in the range images of different echoes are aligned; The coherent accumulation module is used to coherently accumulate M aligned adjacent range images, and intuitively estimate the number of strong scattering points through the peak number and position of the range image after coherent accumulation. and its corresponding distance unit; An energy accumulation module, used to accumulate energy of adjacent high-resolution one-dimensional range images to obtain an energy focusing signal; The detection module uses the Monte Carlo method for detection, uses noise as the reference signal to substitute into the accumulation process of the energy accumulation module, obtains the noise energy focusing signal and the maximum value of the noise energy focusing signal, and according to the given false alarm probability P FA Repeat this step 1 / P FA times, we will get 1 / P FA The maximum value of the noise energy focusing signal is 1 / P FA The maximum values are arranged in descending order, and the value between the first value and the second value in the sorting result is taken as the detection threshold; then the maximum value of the energy focusing signal obtained by the energy accumulation module is compared with the detection threshold, and if the former is greater than the latter, it is judged that there is a target, otherwise it is not; The energy accumulation process of the energy accumulation module includes: Let the i-th high-resolution one-dimensional range image be denoted by R i (f), the i+1th high-resolution one-dimensional range image is denoted as R i+1 (f); According to the length of the target to be detected, a detection window with a length of L is selected, and the length of the detection window is greater than the number of distance units occupied by the target, so that the strong scattering points are completely distributed in one detection window; Sliding windows and accumulating two adjacent one-dimensional range images; the sliding window means that the leading edge position of the detection window is moved to the trailing edge position of the detection window during the previous accumulation during the next accumulation; the accumulation means operating the scattering points in the detection window; if the strong scattering points are distributed in different detection windows, frequency shifting is performed on all the range images to be detected so that all the strong scattering points are distributed in the same detection window, and then sliding windows and accumulating are performed; The accumulation process of the energy accumulation module is: If there is no strong scattering point distribution in the detection window, then conjugate multiply all the distance units of the i-th and i+1-th range images in the detection window and then average them. When the detection window moves to the When the position of a strong scattering point is determined, only the strong scattering points of the i-th and i+1-th range images in the detection window are conjugate multiplied and averaged. This allows the energy at the scattering points to be accumulated.
5. The system for broadband radar detection according to claim 4, characterized in that: When the antenna end receiving processing module receives the next echo, the range image of the next echo and the previous range image repeat the accumulation process of the energy accumulation module and the detection process of the detection module to detect whether there is a target.
