Fractional Fourier Transform Fusion with Pulse - to - Pulse Switching for Low - Slow - Small Target Detection Method

Abstract

The present invention relates to a fractional Fourier transform fusion pulse - to - pulse switching method for low - slow - small target detection, belonging to the field of radar. In the present invention, the echoes with different PRFs are extracted according to the pitch angle pointing, and the echoes with the same pointing are recombined in chronological order into a coherent processing interval, and the echoes with different pointings form multiple CPIs; for a uniformly moving target, the high - Doppler channel uses MTI cascaded with windowed FFT, and the low - channel uses a FIR filter to obtain a range - Doppler map; for a maneuvering target, the fractional Fourier transform is used to obtain the target acceleration, and the quadratic phase term caused by the acceleration is compensated in the echo, and after compensation, MTI cascaded with windowed FFT is used to obtain a range - Doppler map; the obtained range - Doppler map is detected and processed to obtain the angle information of the target. The present invention uses fractional Fourier transform fusion pulse - to - pulse switching to achieve the detection of "low - slow - small" targets, improving the detection probability of "low - slow - small" UAV targets.

Description

Technical Field

[0001] The present invention belongs to the field of radar, and particularly relates to a method for detecting low, slow, and small targets by fusing fractional Fourier transform and inter-pulse switching. Background Art

[0002] In terms of unmanned aerial vehicle (UAV) control, "low, slow, and small" UAVs are characterized by low flight altitude, small size, and slow flight speed, resulting in the inability of existing equipment to detect, identify, and defend in a timely manner. There is no systematic monitoring system and preventive measures, making it difficult to implement real-time monitoring. The phenomenon of "unable to see, unable to control, and unable to strike" often occurs, and the phenomenon of "low, slow, and small" target out of control occurs due to the difficulty in grasping their flight dynamics.

[0003] In terms of target characteristics, small UAVs, as a new type of reconnaissance equipment, have the characteristics of "low, slow, and small". "Low" - the flight altitude is low, and a large amount of ground clutter energy is introduced into the radar beam grazing zone, and the signal-to-clutter ratio becomes the main factor affecting the detection range; "slow" - the flight can be slow to hover, or it can fly intermittently at different speeds without a constant speed, resulting in the divergence of the Doppler spectrum, and it is not easy to detect such targets in the clean area of traditional air defense radars; "small" - the radar cross section (RCS) of the target is small (as small as 0.01 m²), which affects the target detection range. Considering the above factors, small UAVs such as "low, slow, and small" make it difficult for traditional air defense radars to effectively detect targets.

[0004] To solve the problem of detecting "low, slow, and small" targets, a joint detection method of fusing fractional Fourier transform and inter-pulse switching is designed, which comprehensively utilizes the characteristics of energy convergence of fractional Fourier transform and the characteristics of improving Doppler resolution by intra-pulse switching to solve the problem of detecting "slow" targets. Summary of the Invention

[0005] (1) Technical Problems to be Solved

[0006] The technical problem to be solved by the present invention is how to provide a method for detecting low, slow, and small targets by fusing fractional Fourier transform and inter-pulse switching, so as to solve the problem of ineffective detection of "low, slow, and small" targets by traditional ground air defense radars, resulting in missed alarms.

[0007] (2) Technical Solutions

[0008] To solve the above technical problems, the present invention proposes a method for detecting low, slow, and small targets by fusing fractional Fourier transform and inter-pulse switching, and the method includes the following steps:

[0009] S1. The radar starts to work, and the beam control module switches the beam pointing at intervals of the pulse repetition frequency (PRF) in the elevation dimension;

[0010] S2. The radar receives the target echo and performs mixing and filtering processing;

[0011] S3. Perform AD sampling on the filtered signal and perform IQ real-variable complex processing;

[0012] S4. Extract data from the echoes of different PRFs according to the pitch angle pointing, recombine the echoes with the same pointing in chronological order into a coherent processing interval (CPI), and the echoes with different pointings form multiple CPIs;

[0013] S5. Perform pulse compression on the echo data of each CPI;

[0014] S6. For a constant-velocity target, the high channel in the Doppler dimension uses MTI cascaded windowed FFT, and the low channel uses an FIR filter to obtain the range-Doppler map; for a maneuvering target, use the fractional Fourier transform to obtain the target acceleration, compensate for the quadratic phase term caused by the acceleration in the echo, and use MTI cascaded windowed FFT after compensation to obtain the range-Doppler map;

[0015] S7. Perform two-dimensional constant false alarm and clutter map target detection on the obtained range-Doppler map to obtain the sampling unit of the target, the Doppler channel number, and the target amplitude information; perform monopulse angle measurement through the amplitude of the sum-difference channel to obtain the angle information of the target.

[0016] Further, the step S1 specifically includes: taking the PRF as the interval, calculating the beam pointing through the beam control program, and injecting the pointing information into the TR component to lock the phase-coded signal to achieve beam switching.

[0017] Further, in the step S1, the radar radiates a linear frequency modulation signal as follows:

[0018]

[0019]

[0020] x trans (t) is the analytical formula of the transmitted information that changes with time, t is the time, is the transmitted phase information of the i-th radar, f0 is the carrier frequency, B is the signal bandwidth, T is the signal time width, is the pulse initial phase of the i-th radar; the beam control changes the beam pointing in the pitch dimension by controlling the phase change to make the beam point to different angles.

[0021] Further, the step S2 includes: the radar antenna, the transceiver component, and the receiver receive the target echo, which is amplified, filtered by the transceiver component and mixed by the receiver to be converted into an intermediate frequency signal, and the intermediate frequency signal is filtered and amplified.

[0022] Further, in step S2, assuming the distance between the target and the transmitting radar is R1, the received signal form is as follows:

[0023]

[0024]

[0025]

[0026] where c represents the speed of light, R1 is the current distance between the target and the radar; R0 is the initial distance between the target and the radar, v is the target speed, a is the target acceleration, t is the time; f0 is the radar carrier frequency, B is the signal bandwidth, T is the signal time width, is the initial phase.

[0027] Further, step S3 includes:

[0028] S31. The filtered signal is sampled by AD and converted into a digital signal;

[0029] S32. The digital signal is subjected to IQ transformation and converted from a real signal to a complex signal.

[0030] Further, step S6 specifically includes:

[0031] S61. For a uniform target, the high channel in the Doppler dimension is processed by MTI cascaded with windowed FFT, and the low channel uses an FIR filter to obtain the range-Doppler map;

[0032] S62. For a maneuvering target, the echo data in this area is intercepted, and the fractional Fourier transform is used. By rotating the angle in the time-frequency domain and comparing the target energy magnitudes, the magnitude of the rotation angle is determined; according to the obtained rotation angle, according to the formula estimates the target acceleration, where λ is the carrier wavelength, f s is the sampling frequency, T is the signal time length, θ max is the angle corresponding to the transform peak. According to the obtained target acceleration information, the echo quadratic phase term caused by the acceleration is compensated in the target echo a is the acceleration, λ0 is the wavelength corresponding to the center frequency, and then the MTI cascaded with windowed FFT is used to obtain the range-Doppler map.

[0033] Further, in step S6, for a fast maneuvering target, due to the influence of the target acceleration on Doppler spreading, the fractional Fourier transform is used. By rotating the angle in the time-frequency domain and comparing the target energy magnitudes, the magnitude of the rotation angle is determined; the target acceleration is determined according to the rotation angle, and the echo quadratic phase term caused by the acceleration is compensated;

[0034]

[0035] Among them, K p (u, t) is the conversion formula of the fractional Fourier transform in the time-frequency domain, α is the rotation angle in the time-frequency domain, t is time, u is the frequency domain, cot is the cotangent angle, csc is the secant angle, and δ(t) is the impulse function.

[0036] Furthermore, the step S7 includes:

[0037] S71. The Doppler high channel adopts GO-CFAR constant false alarm detection to obtain the sampling unit, Doppler channel number, and amplitude information of the target;

[0038] S72. The Doppler low channel compares the clutter map to obtain the sampling unit and amplitude information of the target;

[0039] S73. Compare the echo energies of the sum channel, azimuth difference channel, and elevation difference channel, compare the sum channel and azimuth difference channel to obtain the azimuth angle of the target, and compare the sum channel and elevation difference channel to obtain the elevation angle of the target.

[0040] Furthermore, in the step S7, the GO-CFAR processing process includes: on the range-Doppler map, M cells are selected on both sides of the range and Doppler dimensions of the target point, the means of these M cells are calculated respectively, the larger one is selected and output, multiplied by the threshold multiplier K as the detection threshold, the detection result of the target amplitude is compared with the detection threshold, the detection result of the target higher than the threshold value is output, and the one lower than the threshold value is judged as fluctuating noise and no result is output.

[0041] (III) Beneficial effects

[0042] The present invention proposes a method for detecting low, slow, and small targets by fusing fractional Fourier transform and pulse-to-pulse switching. Compared with the prior art, when using the method of the present invention for the low, slow, and small target detection algorithm, there are the following two beneficial effects:

[0043] On the one hand, the elevation dimension phase scanning radar adopts beam switching between PRFs. Without changing the elevation coverage range and data rate, the wave dwell time of the CPI is increased by controlling the software control complexity, the resolution of the Doppler dimension is improved, and it is beneficial to the detection of slow targets.

[0044] On the other hand, by using the fractional Fourier transform, time-frequency rotation is performed on the range-Doppler dimension data, and targets flying non-constantly, such as "go-stop-go-stop" of rotor unmanned aerial vehicles, can be detected, the target position and speed information can be obtained, and the target track of the maneuvering flight type unmanned aerial vehicle can be established. Description of the drawings

[0045] Figure 1 The algorithm flow chart adopted by the present invention;

[0046] Figure 2 The beam switching method adopted by the present invention;

[0047] Figure 3 The circuit for realizing constant false alarm rate by selecting the larger value from the average of both sides' units;

[0048] Figure 4 The multi-stage clutter suppression method. Specific implementation manners

[0049] To make the objectives, contents and advantages of the present invention clearer, the following further describes in detail the specific implementation manners of the present invention in conjunction with the accompanying drawings and embodiments.

[0050] When important places face the threat of "low, slow and small" type UAV targets, traditional ground air defense radars face the characteristics of slow flight speed and tortuous flight trajectories of "low, slow and small" type targets. Traditional signal processing methods cannot effectively detect them, resulting in missed alarms. This algorithm comprehensively adopts the fractional Fourier transform and pulse-to-pulse switching to realize the detection of "low, slow and small" type targets, and improves the detection probability of "low, slow and small" type UAV targets.

[0051] To solve the deficiencies of existing ground air defense radars for "slow" targets among "low, slow and small" UAVs: (1) Slow flight speed, and even can hover; (2) Tortuous flight trajectory, sometimes fast and sometimes slow, not flying at a constant speed. Design the joint detection of fractional Fourier transform and pulse-to-pulse switching. The fractional Fourier transform can realize time-frequency fractional detection, re-converge the Doppler spectrum at the fractional order, and extract the target with a tortuous flight; use pulse-to-pulse beam switching to achieve the purpose of improving the frequency domain resolution through fast beam switching and echo recombination, and solve the problem of detecting slow targets.

[0052] The algorithm for detecting "low, slow and small" targets by the fractional Fourier transform and pulse-to-pulse switching adopted by the present invention includes the following steps:

[0053] S1. The radar starts to work, and the beam control module switches the beam pointing at intervals of the pulse repetition frequency PRF in the elevation dimension;

[0054] S2. The radar receives the target echo and performs mixing and filtering processing;

[0055] S3. Perform AD sampling and IQ real-variable complex processing on the filtered signal;

[0056] S4. Extract data from the echoes of different PRFs according to the elevation angle pointing, and recombine the echoes with the same pointing into a coherent processing interval (CPI) in chronological order. The echoes with different pointings form multiple CPIs;

[0057] S5. Perform pulse compression on the echo data of each CPI;

[0058] S6. For a constant-velocity target, the high channel in the Doppler dimension uses MTI cascaded with windowed FFT, and the low channel uses an FIR filter to obtain the range-Doppler map. For a maneuvering target, use the fractional Fourier transform to calculate the target acceleration, compensate for the quadratic phase term caused by the acceleration in the echo, and after compensation, use MTI cascaded with windowed FFT to obtain the range-Doppler map;

[0059] S7. Perform two-dimensional constant false alarm rate and clutter map target detection on the obtained range-Doppler map to obtain information such as the sampling unit of the target, Doppler channel number, and target amplitude. Angle information of the target is obtained by monopulse angle measurement using the amplitudes of the sum and difference channels.

[0060] Among them:

[0061] Step S1 is mainly:

[0062] Taking the PRF as the interval, calculate the beam pointing through the beam control program, and input the pointing information into the TR component to lock the phase code to achieve beam switching;

[0063] Step S2 is mainly:

[0064] The radar antenna, transceiver module, and receiver receive the target echo. After amplification, filtering by the transceiver module, and mixing processing by the receiver, it is converted to an intermediate-frequency signal, and the intermediate-frequency signal is filtered and amplified;

[0065] Step S3 is mainly:

[0066] S31. Sample the filtered signal by AD and convert it into a digital signal;

[0067] S32. Perform IQ transformation on the digital signal to convert it from a real signal to a complex signal;

[0068] Step S4 is mainly:

[0069] Sample and reorganize the sampled digital signal to recombine the echo into CPI data according to the beam pointing;

[0070] Step S5 is mainly:

[0071] Perform pulse compression on the reorganized signal.

[0072] Step S6 mainly includes the following:

[0073] S61. For a uniform target, perform MTI cascaded windowed FFT on the high Doppler channel and use an FIR filter for the low channel to obtain a range-Doppler map;

[0074] S62. For a maneuvering target, intercept the echo data of this area, use fractional Fourier transform, determine the rotation angle by rotating the angle in the time-frequency domain and comparing the target energy. According to the obtained rotation angle, calculate according to the formula, where λ is the carrier wavelength, f s is the sampling frequency, T is the signal time length, θ max is the angle corresponding to the transform peak, estimate the target acceleration, and compensate for the echo quadratic phase term caused by the acceleration in the target echo according to the obtained target acceleration information is the acceleration, λ0 is the wavelength corresponding to the center frequency, and then use MTI cascaded windowed FFT to obtain a range-Doppler map;

[0075] Step S7 mainly includes the following steps:

[0076] S71. Use GO-CFAR constant false alarm detection on the high Doppler channel to obtain information such as the sampling unit, Doppler channel number, and amplitude of the target;

[0077] S72. Compare the clutter map on the low Doppler channel to obtain information such as the sampling unit and amplitude of the target;

[0078] S73. Compare the echo energies of the sum channel, azimuth difference channel, and elevation difference channel, compare the sum channel and azimuth difference channel to obtain the azimuth angle of the target, and compare the sum channel and elevation difference channel to obtain the elevation angle of the target.

[0079] The following further describes the embodiments of the present invention with reference to the accompanying drawings.

[0080] (1) Radar transmission signal

[0081] In step S1, the radar radiates a linear frequency modulation signal as follows:

[0082]

[0083]

[0084] x trans (t) is the analytical formula of the transmission information varying with time, t is the time, is the transmission phase information of the i-th radar, f0 is the carrier frequency, B is the signal bandwidth, T is the signal time width, is the pulse initial phase of the i-th radar.

[0085] The beam control changes the beam pointing in the elevation dimension by controlling the phase change, so that the beam points to different angles.

[0086] (2) Target echo signal

[0087] In step S2, assuming that the distance between the target and the transmitting radar is R1, the received signal form is:

[0088]

[0089]

[0090]

[0091] where c represents the speed of light, R1 is the current distance between the target and the radar; R0 is the initial distance between the target and the radar, v is the target speed, a is the target acceleration, t is the time; f0 is the radar carrier frequency, B is the signal bandwidth, T is the signal time width, is the initial phase.

[0092] As Figure 2 shown, the target echoes are arranged according to the PRF, and different PRF have different elevation pointings. Taking Figure 2 as an example, 5 beams are distributed in the elevation dimension according to the PRF, pointing to 2°, 5°, 8°, 11°, and 14° respectively. Every 5 beams form a cycle. Here, the beams with the same pointing need to be accumulated and recombined into 5 CPIs according to 64 echoes, and rearranged in the order of (1, 6, 11,..., 316), (2, 7, 12,..., 317), (3, 8, 13,..., 318), (4, 9, 14,..., 319), (5, 10, 15,..., 320) to form 5 new CPI data.

[0093] (3) Echo frequency dimension processing

[0094] In step S6, as Figure 3 shown, for a fast and uniform target, the target falls into the high Doppler channel. The method of cascading MTI with windowed FFT can ensure effective clutter suppression performance with a small amount of calculation; for a slow and uniform target, the target falls into the low Doppler channel, and an FIR filter is used to ensure good clutter suppression performance with a small SNR loss.

[0095] For a fast maneuvering target, due to the influence of target acceleration resulting in Doppler spread, the fractional Fourier transform is used. By rotating the angle in the time-frequency domain and comparing the target energy, the rotation angle is determined. The target acceleration is determined according to the rotation angle, and the quadratic phase term of the echo caused by the acceleration is compensated.

[0096]

[0097] Among them, K p (u,t) is the conversion formula of the fractional Fourier transform in the time-frequency domain, α is the rotation angle in the time-frequency domain, t is time, u is the frequency domain, cot is the cotangent angle, csc is the secant angle, and δ(t) is the impulse function.

[0098] (4) Target point extraction

[0099] In step S7, target detection is completed for the range dimension through GO-CFAR constant false alarm detection. As Figure 4 shown is the GO-CFAR processing process: On the range-Doppler map, M cells are selected on each side of the range and Doppler dimensions of the target point, the means of these M cells are calculated respectively, the larger one is selected and output, and it is multiplied by the threshold multiplier K as the detection threshold. The detection result of the target amplitude is compared with the detection threshold. The detection result of the target with an amplitude higher than the threshold value is output, and the one with an amplitude lower than the threshold value is judged as fluctuating noise and no result is output. In engineering applications, usually 16 cells are selected on both sides, and the 2 adjacent cells on the left and right of the cell to be detected do not participate in the statistical generation of the constant false alarm rate threshold of this cell, so as to avoid the influence of the target signal itself (generally, a target may occupy three range cells) on the constant false alarm rate threshold.

[0100] Compared with the prior art, when using the method of the present invention for the low, slow and small target detection algorithm, there are the following two beneficial effects:

[0101] On the one hand, the elevation dimension phase scanning radar adopts beam switching between PRFs. Without changing the elevation coverage range and data rate, the wave dwell time of CPI is increased by controlling the software control complexity, the resolution of the Doppler dimension is improved, which is beneficial to the detection of slow targets.

[0102] On the other hand, by using the fractional Fourier transform, time-frequency rotation is performed on the range-Doppler dimension data, and non-constant speed flying targets such as "go-stop-go-stop" of rotor UAVs can be detected, the target position and speed information can be obtained, and the target track of maneuvering flight UAVs can be established.

[0103] The above is only the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and deformations can be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

Claims

1. A method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching, characterized in that, The method includes the following steps: S1. The radar starts to work, and the beam control module switches the beam pointing at intervals of the pulse repetition frequency (PRF) in the elevation dimension; S2. The radar receives the target echo and performs mixing and filtering processing; S3. Perform AD sampling on the filtered signal and IQ real-variable complex processing; S4. Extract data from the echoes with different PRFs according to the elevation angle pointing, recombine the echoes with the same pointing in chronological order into a coherent processing interval (CPI), and the echoes with different pointings form multiple CPIs; S5. Perform pulse compression on the echo data of each CPI; S6. For a constant-velocity target, the high channel in the Doppler dimension uses MTI cascaded with windowed FFT, and the low channel uses a FIR filter to obtain the range-Doppler map; for a maneuvering target, use the fractional Fourier transform to obtain the target acceleration, compensate the quadratic phase term caused by the acceleration in the echo, and use MTI cascaded with windowed FFT after compensation to obtain the range-Doppler map; S7. Perform two-dimensional constant false alarm and clutter map target detection on the obtained range-Doppler map to obtain the sampling unit, Doppler channel number, and target amplitude information of the target; perform monopulse angle measurement through the amplitude of the sum and difference channels to obtain the angle information of the target.

2. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to claim 1, characterized in that, The specific content of step S1 includes: taking PRF as the interval, calculating the beam pointing through the beam control program, and inputting the pointing information into the TR component to lock the phase-matching code to achieve beam switching.

3. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to claim 2, characterized in that, In step S1, the radar radiates a linear frequency modulation signal as follows: x trans (t) is the analytical formula of the transmitted information varying with time, where t is time, is the transmitted phase information of the i-th radar, f0 is the carrier frequency, B is the signal bandwidth, and T is the signal time width, is the initial pulse phase of the i-th radar; Beam control is achieved by controlling the phase change to alter the beam pointing in the elevation dimension, causing the beam to point at different angles.

4. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to any one of claims 1 - 3, characterized in that, Step S2 includes: the radar antenna, transceiver component, and receiver receive the target echo, and after amplification, filtering by the transceiver component and mixing processing by the receiver, it is converted to an intermediate frequency signal, and the intermediate frequency signal is filtered and amplified.

5. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to claim 4, characterized in that, In step S2, assuming the distance between the target and the transmitting radar is R1, the received signal form is: Where c represents the speed of light, R1 is the current distance of the target from the radar; R0 is the initial distance of the target from the radar, v is the target speed, a is the target acceleration, and t is the time; f0 is the radar carrier frequency, B is the signal bandwidth, T is the signal time width, and is the initial phase.

6. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to claim 4, characterized in that, Step S3 includes: S31. Perform AD sampling on the filtered signal and convert it into a digital signal; S32. Perform IQ transformation on the digital signal to convert it from a real signal to a complex signal.

7. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to claim 6, characterized in that, The specific content of step S6 includes: S61. For a constant-velocity target, in the Doppler dimension, the high channel uses MTI cascaded with windowed FFT, and the low channel uses a FIR filter to obtain the range-Doppler map; S62. For a moving target, the echo data of the intercepted area is subjected to fractional Fourier transform. By rotating the angle in the time-frequency domain and comparing the target energy levels, the magnitude of the rotation angle is determined. Based on the obtained rotation angle, the acceleration of the target is estimated according to the formula, where λ is the carrier wavelength, f s is the sampling frequency, T is the signal time length, θ max is the angle corresponding to the transform peak. Based on the obtained target acceleration information, the quadratic phase term of the echo caused by the acceleration is compensated in the target echo where a is the acceleration and λ0 is the wavelength corresponding to the center frequency. Then, MTI cascaded with windowed FFT is used to obtain the range-Doppler map.

8. The method for detecting low, slow and small targets by fusing fractional Fourier transform with pulse - to - pulse switching according to claim 6, characterized in that, In step S6, for a fast maneuvering target, due to the influence of target acceleration resulting in Doppler spread, use the fractional Fourier transform, determine the rotation angle size by rotating the angle in the time-frequency domain and comparing the target energy size; determine the target acceleration according to the rotation angle and compensate the quadratic phase term of the echo caused by the acceleration; Among them, K p (u,t) is the conversion formula of the fractional Fourier transform in the time-frequency domain, α is the rotation angle in the time-frequency domain, t is time, u is the frequency domain, cot is the cotangent angle, csc is the secant angle, and δ(t) is the impulse function.

9. The fractional Fourier transform fusion pulse - to - pulse switching method for detecting low - slow - small targets according to claim 7, wherein, Step S7 includes: S71. The high Doppler channel uses GO-CFAR constant false alarm detection to obtain the sampling unit, Doppler channel number, and amplitude information of the target; S72. The low Doppler channel compares the clutter map to obtain the sampling unit and amplitude information of the target; S73. Compare the echo energies of the sum channel, azimuth difference channel, and elevation difference channel. By comparing the sum channel and azimuth difference channel, obtain the azimuth angle of the target. By comparing the sum channel and elevation difference channel, obtain the elevation angle of the target.

10. The fractional Fourier transform fusion pulse - to - pulse switching method for detecting low - slow - small targets according to claim 9, wherein, In step S7, the GO-CFAR processing procedure includes: On the range-Doppler map, select M cells on each side of the range and Doppler dimensions of the target point, calculate the mean values of these M cells respectively, select the larger one of the two and output it, multiply it by the threshold multiplier K as the detection threshold, compare the detection result of the target amplitude with the detection threshold, output the detection result of the target if it is higher than the threshold value, and judge it as fluctuating noise if it is lower than the threshold value and do not output the result.

Images