Multi-dimensional domain interference suppression method based on LFMCW radar echo characteristics

Abstract

The invention relates to the technical field of linear frequency modulation continuous wave (LFMCW) radar signal processing, and discloses a multi-dimensional domain interference suppression method based on LFMCW radar echo characteristics, and the method mainly comprises the steps: firstly, carrying out the weighted beam forming of an array receiving signal in a space domain, so as to reduce the interference influence in an unexpected direction; performing short-range gain control filtering and time domain compression in a fast time domain; moving target display, moving target detection and adaptive motion clutter elimination are carried out in the Doppler frequency domain; and then carrying out improved variable index constant false alarm detection in an energy domain, reducing the influence of energy fluctuation interference, and finally outputting a multi-dimensional domain interference suppression result. According to the method, the interference is effectively suppressed in the airspace, the fast time domain, the Doppler frequency domain and the energy domain through the characteristics of the LFMCW radar echo signal, on one hand, the problem of overload of a radar receiver caused by strong interference can be avoided, and on the other hand, the target detection capability and the false alarm resistance of the radar in a complex environment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of LFMCW radar signal processing, and in particular to a multi-dimensional domain interference suppression method for LFMCW radar. Background Art

[0002] The potential interference of LFMCW radar widely exists in the airspace, fast time domain, Doppler frequency domain and energy domain. The airspace interference enters the receiving system through the sidelobe of the receiving antenna beam. Strong interference in the near area will cause receiver overload. Static and moving interference will cause false targets to be generated, and the energy fluctuation of the interference signal is large. Therefore, it is necessary to suppress the multi-dimensional interference to improve the LFMCW radar's anti-false alarm capability and target detection performance, and avoid receiver overload.

[0003] The current interference suppression methods mainly focus on space-time two-dimensional processing, but it is still difficult to deal with near-field interference entering from the main lobe of the beam, especially for LFMCW radar, which can easily cause receiver overload. Therefore, it is necessary to adopt more dimensional interference suppression technology based on the echo characteristics of LFMCW radar.

[0004] For spatial interference, a low sidelobe beam can be obtained through a weighted beamforming algorithm, thereby suppressing interference signals in undesired directions. For near-zone and static and moving clutter interference, according to the time-frequency characteristics of the LFMCW radar echo, near-zone interference is positively correlated with time, and can be suppressed by a high-pass filter. Static and moving clutter interference is related to the Doppler frequency domain, and the Doppler frequency of the interference signal can be separated by an adaptive moving target detection algorithm, and the interference Doppler frequency domain information can be estimated to adaptively suppress the interference area. For interference with amplitude fluctuations, adaptive detection and suppression can be performed through constant false alarm detection. Summary of the invention

[0005] The object of the present invention is to provide a multi-dimensional domain interference suppression method based on LFMCW radar echo characteristics, so as to overcome the problems existing in the prior art.

[0006] The technical solution to achieve the purpose of the present invention is: a multi-dimensional domain interference suppression method based on LFMCW radar echo characteristics, comprising the following steps:

[0007] 1. A multi-dimensional domain interference suppression method based on LFMCW radar echo characteristics, characterized in that it includes the following steps:

[0008] Step S100, performing weighted beamforming on the interference echo signal received by the array antenna, and outputting a signal after spatial interference suppression;

[0009] Step S200, the input signal after spatial interference suppression is subjected to a short-range gain control filter to perform short-range interference suppression, and a filtered difference frequency interference signal is output;

[0010] Step S300, using N-point fast Fourier transform to perform time domain compression on the filtered difference frequency echo signal, and output the interference signal after time domain compression;

[0011] Step S400, displaying moving targets on the time-domain compressed input interference signal along the Doppler dimension, and outputting a signal in which static interference is filtered out;

[0012] Step S500, using M-point fast Fourier transform to perform moving target detection on the input signal after filtering out static interference, and outputting Doppler frequency separated interference signal;

[0013] Step S600, performing adaptive motion clutter elimination on the interference signal separated by Doppler frequency, estimating the interference quantity, Doppler center frequency and spectrum width, performing Doppler frequency domain suppression on the interference area signal, and outputting the interference signal after Doppler frequency domain suppression;

[0014] Step S700, perform improved variable index constant false alarm detection on the input interference signal after Doppler frequency domain suppression, and compare the reference point amplitude value with the left and right sliding window amplitude mean estimation value in turn according to the interference background adaptive selection algorithm, suppress the interference signal with large energy fluctuation, and finally output the multi-dimensional domain interference suppression result.

[0015] 2. The weighted beam forming in step S100 includes:

[0016] Step S101: for a receiving array of P elements, the received signal steering vector can be expressed as a(θ)=[1, e -j2πdsinθ / λ ,…,e -j2π(P-1)dsinθ / λ ] T , the desired direction is θ0, and the conventional beamforming weight coefficient vector is calculated as ω CBF =a(θ0), the selected windowing coefficient vector is ω e , and the weighted coefficient is ω WBF =ω e ω CBF ;

[0017] Step S102: Perform weighted beamforming on the received signal x(t) and output a signal after spatial interference suppression.

[0018] 3. The short-range gain control filtering in step S200 includes:

[0019] Step S201: Combine the input spatial interference suppression signal with the local oscillator signal s t (t) Mixing is performed to obtain the difference frequency echo signal

[0020] Step S202, perform short-range gain control filtering on the difference frequency echo signal. The time domain expression of the high-pass filter is h s (t), output the filtered difference frequency echo signal s hb (t) = s b (t)*h s (t).

[0021] 4. The step S400 of displaying the moving target includes:

[0022] Step S401, performing an N×M dimensional matrix transposition operation on the fast time dimension data after time domain compression to obtain data output along the Doppler dimension;

[0023] Step S402, accumulating and averaging the input data along the Doppler dimension, and then subtracting the input data from the average value in sequence to remove the DC component, and outputting a signal with static ground interference removed;

[0024] 5. The step S600 of adaptive motion clutter elimination includes:

[0025] Step S601, sorting the interference signals separated by Doppler frequency, and selecting the largest Q interference data points;

[0026] Step S602, performing Doppler index association on the input Q maximum interference data points, and outputting the Doppler index associated data points;

[0027] Step S603, calculating the average value and upper and lower limits of the Doppler index of the input associated data points, and outputting the number of interference signals, Doppler frequency and spectrum width;

[0028] Step S604, storing and updating the input interference signal quantity, Doppler frequency and spectrum width information, and outputting updated interference area information;

[0029] Step S605, suppress the interference signal separated by Doppler frequency according to the input interference area information, if the input signal Doppler index is in the interference area, reduce the signal amplitude by K times, otherwise the signal amplitude remains unchanged, and output the interference signal after Doppler frequency domain suppression.

[0030] 6. The step S700 of improving variable index constant false alarm detection includes:

[0031] Step S701: The interference signal after Doppler frequency domain suppression currently input is used as the reference unit x d , calculate the mean of the D points before and after and And output.

[0032] Step S702: Calculate the change index V of the front and rear sliding windows respectively according to the input mean value VI And the mean ratio of the leading and trailing edges V MR , and output the change index and the mean ratio of the front and back edges. The change index and the mean ratio of the front and back edges are calculated as follows:

[0033]

[0034] Step S703: compare the input change index and the front and rear edge mean ratio with the threshold value K VI and K MR Compare, if V VI ≤K VI It can be judged as uniform clutter if It can be judged that the mean values ​​are the same, and the clutter environment judgment result is output.

[0035] Step S704, according to the clutter environment judgment result, select the CFAR algorithm to obtain the current detection threshold value, and output the detection threshold value;

[0036] Step S705 , comparing the reference unit with the detection threshold value, suppressing interference signals with large energy fluctuations, and outputting a multi-dimensional domain interference suppression result.

[0037] Compared with the prior art, the present invention has the following significant advantages:

[0038] The present invention makes full use of the echo time-frequency characteristics of the LFMCW radar, performs multi-dimensional domain suppression on spatial interference, near-zone interference, fluctuation interference, and static and moving clutter interference, avoids the receiver overload problem caused by interference, and improves the radar system's false alarm resistance and target detection capabilities.

[0039] Furthermore, the multi-dimensional domain interference suppression algorithm adopted by the present invention has small computational complexity and is easy to implement in engineering. The method can be introduced on the basis of the LFMCW radar target detection algorithm to achieve interference suppression. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Flow chart of the multi-dimensional interference suppression method based on LFMCW radar echo characteristics of the present invention

[0041] Figure 2 Antenna pattern simulation results for the weighted beamforming algorithm

[0042] Figure 3 This is the curve diagram of short-range gain control filtering and near-range interference change

[0043] Figure 4 Comparison chart of static and dynamic clutter interference suppression

[0044] Figure 5To improve the variable index constant false alarm detection result diagram DETAILED DESCRIPTION

[0045] To facilitate understanding of the present invention, the method of the present invention is further described below in conjunction with the accompanying drawings and embodiments. The present invention can be implemented in different forms and is not limited to the embodiments described in the text. On the contrary, the embodiments are provided to make the content disclosed in the present invention more thorough and comprehensive. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0046] The basic scheme of the present invention includes the following 7 steps, and the specific process of the method is as follows: Figure 1 shown.

[0047] For a receiving array with P = 16 elements, the desired direction is 0°, and the received signal steering vector can be expressed as a(θ) = [1, e -j2πdsinθ / λ ,…,e -j(P-1)πdsinθ / λ ] T , the incident interference signal angle is 10°, and the interference signal is an LFM signal with a time delay τ(t)

[0048] s(t)=A0exp(j[2πf0(t-τ(t))+πμ(t-τ(t)) 2 ]),

[0049] Among them, A0 is the amplitude of the received echo signal; f0 is the initial frequency of the transmitted signal frequency modulation; μ is the frequency modulation slope. Then the array received signal is x(t) = a(10°)s(t).

[0050] In step S100, the conventional beamforming weight coefficient vector is ω CBF =a(0°), the weighted window function selects the Chebyshev window ω chb , the weighting coefficient is ω WBF =ω e a(0°). The array receives the signal and performs weighted beamforming. Since the interference signal is in the beam sidelobe area at this time, the interference signal amplitude is greatly reduced after spatial domain suppression, and the signal form does not change. The signal after spatial domain interference suppression is

[0051]

[0052] Among them, ɑ is the suppression factor, which is related to the beam sidelobe level.

[0053] Step S200: The input spatial interference suppression signal is combined with the local oscillator signal s i (t) Mixing is performed to obtain the difference frequency interference signal Where β is the amplitude factor, f bis the time domain delay frequency, is the phase change. The difference frequency echo signal is subjected to short-range gain control filtering, and the time domain expression of the high-pass filter is h s (t), at this time, the near-field component in the interference signal is suppressed, and the difference frequency echo signal after filtering is

[0054]

[0055] Among them, f b1 is the residual mid- and far-range time domain delayed frequency component.

[0056] Step S300, using N-point fast Fourier transform to transform M frequency sweep cycles T c The difference frequency echo signal after inner filtering is compressed in time domain, and the interference signal after time domain compression is output as

[0057]

[0058] Among them, T s is the sampling period, f d is the interference signal Doppler frequency, k=0,1,2,...N-1.

[0059] Step S400, performing an N×M dimensional matrix transposition operation on the fast time dimension data after time domain compression to obtain data S output along the Doppler dimension b (m, k). The input data along the Doppler dimension are accumulated and averaged to obtain the mean value

[0060]

[0061] Then the input data is subtracted from the mean value in turn to remove the DC component and suppress the static interference. The signal after suppressing the static interference is:

[0062]

[0063] Step S500, using M-point fast Fourier transform to perform moving target detection on the input signal after filtering out static interference, and obtain the interference signal separated by Doppler frequency as

[0064]

[0065] Among them, f d1 To suppress the Doppler frequency of the interference signal remaining after zero frequency, l = 0, 1, 2, ...M-1.

[0066] Step S600, first sort the interference signals separated by Doppler frequency, and select the largest Q=8 interference data points as

[0067] S bz (r, k) = max(Sbz (l, k), Q)

[0068] Wherein, r=0, 1, 2, ... Q-1, is the index of the maximum Q interference data points;

[0069] Secondly, the Doppler index of the input Q maximum interference data points is traversed in a loop. If the indexes of multiple data points are less than the correlation threshold, they are considered to be the same interference and merged into the same data point data, and the data point after Doppler index correlation is output;

[0070] Again, the Doppler indices of the associated data points are averaged As well as the upper and lower limits Δr, the Doppler frequency and spectrum width of the interference signal are calculated according to the radar related parameters, and the number of interference signals is output Doppler frequency f dz and the spectral width Δf dz ;

[0071] Then the input interference signal quantity, Doppler frequency and spectrum width information are stored and updated, and the updated interference area information is output;

[0072] Finally, the interference signal separated by Doppler frequency is suppressed according to the input interference area information. If the input signal Doppler index is in the interference area, the signal amplitude is reduced by K=4 times, otherwise the signal amplitude remains unchanged, and the interference signal after Doppler frequency domain suppression is output.

[0073] Step S700, the false alarm probability is set to P fa =10 -4 , the left and right sliding window size is D = 8. First, the interference signal after the Doppler frequency domain suppression of the current input is used as the reference unit x d , calculate the mean of the D points before and after and And output;

[0074] Secondly, calculate the change index V of the front and rear sliding windows according to the input mean VI And the mean ratio of the leading and trailing edges V MR , and output the change index and the mean ratio of the front and back edges. The change index and the mean ratio of the front and back edges are calculated as follows:

[0075]

[0076] Again, the input change index and the mean ratio of the leading and trailing edges are compared with the threshold value K VI =4.36 and K MR =1.806, if V VI ≤K VI It can be judged as uniform clutter if It can be judged that the mean values ​​are the same, and the clutter environment judgment result is output;

[0077] Then, according to the result of the clutter environment judgment, the optimal CFAR algorithm is selected to obtain the detection threshold value. If it is judged to be uniform interference, CA-CFAR is selected. If it is judged to be an interference edge environment, GO-CFAR is selected. If there is interference in both the front and rear windows, TM-CFAR is selected and the detection threshold value V is output. T =T CFAR ·Z CFAR ;

[0078] Finally, the reference unit is compared with the detection threshold value to suppress interference signals with large energy fluctuations, and the multi-dimensional domain interference suppression result is output.

Claims

1. A multi-dimensional domain interference suppression method based on LFMCW radar echo characteristics, characterized in that: The following steps are involved: Step S100, performing weighted beamforming on the interference echo signal received by the array antenna, and outputting a signal after spatial interference suppression; Step S200, the input signal after spatial interference suppression is subjected to a short-range gain control filter to perform short-range interference suppression, and a filtered difference frequency interference signal is output; Step S300, using N-point fast Fourier transform to perform time domain compression on the filtered difference frequency echo signal, and output the interference signal after time domain compression; Step S400, displaying moving targets on the time-domain compressed input interference signal along the Doppler dimension, and outputting a signal in which static interference is filtered out; Step S500, using M-point fast Fourier transform to perform moving target detection on the input signal after filtering out static interference, and outputting Doppler frequency separated interference signal; Step S600, performing adaptive motion clutter elimination on the interference signal separated by Doppler frequency, estimating the interference quantity, Doppler center frequency and spectrum width, performing Doppler frequency domain suppression on the interference area signal, and outputting the interference signal after Doppler frequency domain suppression; Step S700, perform improved variable index constant false alarm detection on the input interference signal after Doppler frequency domain suppression, and compare the reference point amplitude value with the left and right sliding window amplitude mean estimation value in turn according to the interference background adaptive selection algorithm, suppress the interference signal with large energy fluctuation, and finally output the multi-dimensional domain interference suppression result.

2. The multi-dimensional interference suppression method based on LFMCW radar according to claim 1 is characterized in that: The weighted beam forming in step S100 includes: Step S101: for a receiving array of P elements, the received signal steering vector can be expressed as a(θ)=[1,e -j2πd sinθ / λ ,…,e -j2x(P-1)d sinθ / λ ] T , the desired direction is θ0, and the conventional beamforming weight coefficient vector is calculated as ω CBF =a(θ0), the selected windowing coefficient vector is ω e , and the weighted coefficient is ω WBF =ω e ω CBF ; Step S102: Perform weighted beamforming on the received signal x(t) and output a signal after spatial interference suppression.

3. The multi-dimensional interference suppression method based on LFMCW radar according to claim 1 is characterized in that: The step S200 of short-range gain control filtering includes: Step S201: Combine the input spatial interference suppression signal with the local oscillator signal s t (t) Mixing is performed to obtain the difference frequency echo signal Step S202, perform short-range gain control filtering on the difference frequency echo signal. The time domain expression of the high-pass filter is h s (t), output the filtered difference frequency echo signal s hb (t) = s b (t)*h s (t).

4. The multi-dimensional interference suppression method based on LFMCW radar according to claim 1 is characterized in that: The step S400 of displaying the moving target includes: Step S401, performing an N×M dimensional matrix transposition operation on the fast time dimension data after time domain compression to obtain data output along the Doppler dimension; Step S402: accumulate and average the input data along the Doppler dimension, then subtract the input data from the average value in sequence to remove the DC component, and output a signal with static ground interference removed.

5. The multi-dimensional interference suppression method based on LFMCW radar according to claim 1 is characterized in that: The step S600 of adaptive motion clutter elimination includes: Step S601, sorting the interference signals separated by Doppler frequency, and selecting the largest Q interference data points; Step S602, performing Doppler index association on the input Q maximum interference data points, and outputting the Doppler index associated data points; Step S603, calculating the average value and upper and lower limits of the Doppler index of the input associated data points, and outputting the number of interference signals, Doppler frequency and spectrum width; Step S604, storing and updating the input interference signal quantity, Doppler frequency and spectrum width information, and outputting updated interference area information; Step S605, suppress the interference signal separated by Doppler frequency according to the input interference area information, if the input signal Doppler index is in the interference area, reduce the signal amplitude by K times, otherwise the signal amplitude remains unchanged, and output the interference signal after Doppler frequency domain suppression.

6. The multi-dimensional interference suppression method based on LFMCW radar according to claim 1 is characterized in that: The step S700 of improving variable index constant false alarm detection includes: Step S701: The interference signal after Doppler frequency domain suppression currently input is used as the reference unit x d , calculate the mean of the D points before and after and And output; Step S702: Calculate the change index V of the front and rear sliding windows respectively according to the input mean value VI And the mean ratio of the leading and trailing edges V MR , and output the change index and the mean ratio of the front and back edges. The change index and the mean ratio of the front and back edges are calculated as follows: Step S703: compare the input change index and the front and rear edge mean ratio with the threshold value K VI and K MR Compare, if V VI ≤K VI It can be judged as uniform clutter if It can be judged that the mean values ​​are the same, and the clutter environment judgment result is output; Step S704, according to the clutter environment judgment result, select the CFAR algorithm to obtain the current detection threshold value, and output the detection threshold value; Step S705 , comparing the reference unit with the detection threshold value, suppressing interference signals with large energy fluctuations, and outputting a multi-dimensional domain interference suppression result.

Images