Method for improving robustness of sparse array of millimeter wave radar

Abstract

The invention discloses a method for improving the robustness of a sparse array of a millimeter wave radar. According to the method, under the condition that the aperture of the array is as large as possible, array holes caused by sparse arrays are supplemented in a mode of transmitting AB waves, so that grating lobes are eliminated, and the ratio of main lobes to side lobes is increased. Compared with the prior art, the information amount obtained by the array can be increased, and the hole positions of the sparse array can be supplemented without errors.

Description

Technical Field

[0001] The present invention relates to a method for eliminating grating lobes of a sparse array of millimeter-wave radars and improving the main-to-side lobe ratio. Background Art

[0002] Currently, the autonomous driving industry is booming. As an important sensing device, millimeter-wave radar is an important part of intelligent driving. The angle measurement performance of the radar directly affects the sensing performance. Currently, in-vehicle millimeter-wave radars have high requirements for the angular resolution and accuracy of the radar. Therefore, sparse arrays are used in radars to expand the aperture as much as possible, and the sparsity of the array leads to higher side lobes and even grating lobes.

[0003] In reference [1], two sub-arrays are formed, and the array is expanded inward according to the covariance matrix of the received signals, which solves the problem of poor robustness caused by the sparsity of the array. In reference [2], the array aperture is increased by a conjugate virtual array, and the increased aperture is used for decorrelation. Neither method increases the information volume, and there are errors in the supplementary virtual holes in the case of multiple targets, which may lead to angle measurement errors with a certain probability.

[0004] References

[0005] [1] Mao Weining, Qian Jin, Chen Jianrun. A method for expanding a linear array: CN202010708433.8[P]. CN111736118A[2024-11-26].

[0006] [2] Sun Zhiguo, Bai Qiaosen, Bai Yongzhen, et al. Improved weighted spatial smoothing algorithm based on virtual array expansion[J]. Systems Engineering and Electronics, 2023, 45(1): 250-256. DOI: 10.12305 / j.issn.1001-506X.2023.01.29. Summary of the Invention

[0007] The object of the present invention is to expand the virtual aperture of the array, supplement the holes in the sparse array, improve the main-to-side lobe ratio during angle measurement, and further increase the robustness of angle measurement without adding physical array elements.

[0008] To achieve the above object, the technical solution of the present invention discloses a method for improving the robustness of a sparse array of millimeter-wave radars, which is characterized by including the following steps:

[0009] Step 1, configure the positions of each radar based on a pre-designed sparse array D, select a target spacing value from the spacing values between adjacent two elements in all sparse arrays D according to the array aperture to be supplemented, and further determine a proportionality coefficient k based on the target spacing value;

[0010] Step 2: Set the center frequencies of the two radar waveform parameters based on the proportionality coefficient k to complete the design of the linear frequency modulation continuous wave waveform. Among them, the center frequency of the first radar waveform parameter is f1, and the center frequency of the second radar waveform parameter is f2, where f2 = k × f1;

[0011] Step 3: Based on the set radar waveform parameters, make each radar continuously transmit two radar waveforms with different center frequencies;

[0012] Step 4: Perform ADC sampling on the radar echo;

[0013] Step 5: Perform two FFTs on the rows of the ADC data matrix of the radar waveform with the center frequency f1 obtained in Step 4 to obtain 2DFFT_1 data;

[0014] Perform two FFTs on the rows of the ADC data matrix of the radar waveform with the center frequency f2 obtained in Step 4 to obtain 2DFFT_2 data;

[0015] Step 6: Incoherently accumulate the 2DFFT_1 data and the 2DFFT_2 data to obtain the position in the target matrix, and obtain the data 2DFFT_1(rIdx, dIdx) in the 2DFFT_1 data and the data 2DFFT_2(rIdx, dIdx) in the 2DFFT_2 data, where rIdx is the abscissa and dIdx is the ordinate;

[0016] Step 7: After zeroing the phase of the first channel of the data 2DFFT_1(rIdx, dIdx) and the data 2DFFT_2(rIdx, dIdx) obtained in the previous step, arrange them in order into a set of data x, then the obtained data x realizes the supplement of the array aperture.

[0017] Preferably, in Step 1, let the target spacing value be represented as Δd, then the proportionality coefficient k = 1 / Δd.

[0018] Preferably, in Step 2, except for the center frequency, the other parameters of the two radar waveform parameters are the same.

[0019] Preferably, after Step 4, record the ADC sampling data in a memory. Then, the first half of the data recorded in the memory is the ADC data matrix of the radar waveform with the center frequency f1, and the second half of the data is the ADC data matrix of the radar waveform with the center frequency f2.

[0020] Preferably, the set of data x obtained in Step 7 is expressed as:

[0021] x = [aBS(x1)*exp(-1j*angle(x1), ABS(x2)*exp(-1j*angle(x2))

[0022] In the formula: ABS() is to take the absolute value, angle() is to take the phase, x1 is the data 2DFFT_1(rIdx, dIdx), which is the echo signal corresponding to the radar waveform with the center frequency f1, and there is:

[0023]

[0024] where D1 = D, lmd1 is the wavelength of the radar waveform with the center frequency f1, and θ is the angle;

[0025] x2 is the data 2DFFT_2(rIdx, dIdx), which is the echo signal corresponding to the radar waveform with the center frequency f2, and there is:

[0026]

[0027] where D2 = k × D1, lmd2 is the wavelength of the radar waveform with the center frequency f2, and lmd2 = k × lmd1.

[0028] Preferably, after the step 7, it further includes:

[0029] Step 8, performing DBF angle measurement based on the data x, there is:

[0030] θ e = argmax(A(θ)x)

[0031] where θ e is the angle estimation value, argmax(·) is the value of θ when the vector takes the maximum value, is the steering vector, and d ∈ [D1 ∪ D2].

[0032] The present invention proposes a method that can, when the array aperture is as large as possible, supplement the array holes caused by array sparsity by emitting AB waves, thereby eliminating grating lobes and improving the main-to-side lobe ratio. Compared with the prior art solutions, the present invention can increase the amount of information obtained by the array and accurately supplement the hole positions of the sparse array. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is the flowchart of the present invention;

[0034] Figure 2 is the layout schematic diagram;

[0035] Figure 3 shows the angle measurement spatial spectrum when only a single waveform is used;

[0036] Figure 4 shows the angle measurement spatial spectrum using the AB wave method proposed by the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0037] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of this application.

[0038] As Figure 1 shown, an embodiment of the present invention discloses a method for improving the robustness of a millimeter-wave radar sparse array, which uses a linear frequency modulation continuous wave to perform phase compensation, rearrangement, and angle measurement on echo data, including the following steps:

[0039] Step S1: Configure the positions of each radar based on the pre-designed sparse array D. Let the sparse array D be expressed as [d1, d2,..., d n ,..., d N , where d n is the position of the nth radar. Calculate the spacing values between adjacent two elements in the sparse array D, and a total of N - 1 spacing values are obtained. Select the target spacing value Δd from all the spacing values according to the array aperture to be supplemented.

[0040] Set two sets of MIMO radar waveform parameters to complete the linear frequency modulation continuous wave waveform design, including the number of sampling points, sampling rate, number of waveforms, waveform period, frequency modulation slope, etc. In the embodiment of the present invention, the center frequencies of the two sets of MIMO radar waveform parameters are different. The center frequency of the first set of MIMO radar waveform parameters is f1, and the center frequency of the second set of MIMO radar waveform parameters is f2. Other parameters are the same, and f2 = k × f1, where k = 1 / Δd. Based on the set parameters, continuously transmit two radar waveforms with different center frequencies, that is, each frame contains two radar waveforms with different center frequencies.

[0041] For example, the designed sparse array D is [0, 2, 4, 6, 8]. At this time, the spacing values between adjacent two elements are all 2, and the array apertures between element 0 and element 2, and between element 2 and element 4 in the sparse array D are supplemented. Then the target spacing value Δd = 2, and further calculated k = 0.5. Adjust the center frequency f2 so that f2 = 0.5 × f1.

[0042] In the above case, the elements in the sparse array D form an arithmetic progression. However, in other cases, the elements in the sparse array D do not form an arithmetic progression. For example, if the designed sparse array D is [0, 3, 4, 6, 7], the spacing values between adjacent elements are 3, 1, 2, 1. At this time: If it is necessary to supplement the array aperture between element 0 and element 3 in the sparse array D, the target spacing value Δd = 3, k = 1 / 3, and the center frequency f2 is adjusted such that f2 = 1 / 3 × f1; If it is necessary to supplement the array aperture between element 4 and element 6 in the sparse array D, the target spacing value Δd = 2, and the center frequency f2 is adjusted such that f2 = 0.5 × f1.

[0043] Step S2: Perform ADC sampling on the current system and record it in a memory. Since two radar waveforms are continuously transmitted in step S1, the first half of the data recorded in the memory is the ADC data matrix of the radar waveform with the center frequency f1, and the second half of the data is the ADC data matrix of the radar waveform with the center frequency f2.

[0044] Step S3: Perform two FFTs on the rows of the ADC data matrix of the radar waveform with the center frequency f1 to obtain 2DFFT_1 data.

[0045] Step S4: Perform two FFTs on the rows of the ADC data matrix of the radar waveform with the center frequency f2 to obtain 2DFFT_2 data.

[0046] Step S5: Incoherently accumulate the 2DFFT_1 data and the 2DFFT_2 data to obtain the position in the target matrix, with the abscissa being rIdx and the ordinate being dIdx.

[0047] Step S6: Set the phase of the first channel of 2DFFT_1(rIdx, dIdx) in the matrix data of 2DFFT_1 to zero, and then set the phase of the first channel of 2DFFT_2(rIdx, dIdx) in the matrix data of 2DFFT_2 to zero, and arrange the two sets of zero-phase data in order to form a set of data x. Then:

[0048] x = [ABS(x1)*exp(-1j*angle(x1), ABS(x2)*exp(-1j*angle(x2))

[0049] In the formula: x1 is the echo signal corresponding to the radar waveform with the center frequency f1, and there is:

[0050]

[0051] where D1 = D, lmd1 is the wavelength of the radar waveform with the center frequency f1, and θ is the angle;

[0052] x2 is the echo signal corresponding to the radar waveform with a center frequency of f2, and there is:

[0053]

[0054] Among them, D2 = k × D1, λd2 is the wavelength of the radar waveform with a center frequency of f2, and λd2 = k × λd1;

[0055] x1 is the data 2DFFT_1(rIdx, dIdx) in 2DFFT_1;

[0056] x2 is the data 2DFFT_2(rIdx, dIdx) in 2DFFT_2.

[0057] In the embodiment of the present invention, assuming that the sparse array D is [0, 2, 4, 6, 8], the array apertures between element 0 and element 2, and between element 2 and element 4 in the sparse array D are supplemented, then k = 0.5, and D2 = k × D1, which is [0, 1, 2, 3, 4]. Therefore, the obtained data Among them, d ∈ [D1 ∪ D2]. Therefore, the array aperture is supplemented.

[0058] Step S7: Perform DBF angle measurement based on the data x, and there is:

[0059] θ e = argmax(A(θ)x)

[0060] Among them, θ e is the angle estimation value, and argmax(·) is the value of θ when the vector takes the maximum value, is the steering vector.

[0061] Set the element spacing to [0.0019, 0.0039, 0.0097, 0.0117, 0.0136, 0.0156, 0.0175, 0.0195, 0.0253, 0.0273, 0.0292, 0.0312, 0.0331, 0.0351, 0.0410, 0.0429, 0.0449, 0.0468, 0.0487, 0.0507, 0.0565, 0.0585, 0.0604, 0.0624], with a total of 24 elements, and the unit is m. This column is a sparse array, and the array layout is as Figure 2 shown, where the blue ones are physical elements and the red ones are holes. Due to the holes, the main side lobe is relatively large and the robustness is poor.

[0062] Set two signal sources. The distances, speeds, and gains of the two signals are equal. The angle of signal 1 is 1.3°, and the angle of signal 2 is 4.8°. Figure 3 is the angle measurement spatial spectrum when only using a single waveform,Figure 4 For the AB wave method (the center frequency of wave A f1 is 78 GHz, and the center frequency of wave B f2 is 39 GHz) proposed by the present invention.

[0063] It can be observed that due to the holes, the main-to-side lobe ratio of a single waveform is -8 dB. After the holes are supplemented by the method proposed by the present invention, the main-to-side lobe ratio of the array is -10.4 dB, proving that the robustness of the array is improved.

Claims

1. A method for improving the robustness of a millimeter-wave radar sparse array, characterized in that Including the following steps: Step 1: Configure the positions of each radar based on a pre-designed sparse array D, select a target spacing value from the spacing values between adjacent elements of all sparse arrays D according to the array aperture to be supplemented, and further determine a proportionality coefficient k based on the target spacing value; Step 2: Set the center frequencies of two radar waveform parameters based on the proportionality coefficient k to complete the design of the linear frequency modulation continuous wave waveform. Among them, the center frequency of the first radar waveform parameter is f1, the center frequency of the second radar waveform parameter is f2, and f2 = k×f1; Step 3: Based on the set radar waveform parameters, make each radar continuously transmit two radar waveforms with different center frequencies; Step 4: Perform ADC sampling on the radar echo; Step 5: Perform two FFTs on the rows of the radar waveform ADC data matrix with a center frequency of f1 obtained in Step 4 to obtain 2DFFT_1 data; Perform two FFTs on the rows of the radar waveform ADC data matrix with a center frequency of f2 obtained in Step 4 to obtain 2DFFT_2 data; Step 6: Incoherently accumulate the 2DFFT_1 data and the 2DFFT_2 data to obtain the position in the target matrix, and obtain the data 2DFFT_1(rIdx,dIdx) in the 2DFFT_1 data and the data 2DFFT_2(rIdx,dIdx) in the 2DFFT_2 data, where rIdx is the abscissa and dIdx is the ordinate; Step 7: After zeroing the phase of the first channel of the data 2DFFT_1(rIdx,dIdx) and the data 2DFFT_2(rIdx,dIdx) obtained in the previous step, arrange them in order into a set of data x, then the obtained data x realizes the supplementation of the array aperture.

2. A method for improving the robustness of a sparse array of millimeter-wave radars according to claim 1, characterized in that, In Step 1, assuming the target spacing value is represented as Δd, then the proportionality coefficient k = 1 / Δd.

3. A method for improving the robustness of a sparse array of millimeter-wave radars according to claim 1, characterized in that In Step 2, except for the center frequency, the other parameters of the two radar waveform parameters are the same.

4. A method for improving the robustness of a sparse array of millimeter-wave radars according to claim 1, characterized in that, After Step 4, record the ADC sampling data in a memory. Then, the first half of the data recorded in the memory is the radar waveform ADC data matrix with a center frequency of f1, and the second half of the data is the radar waveform ADC data matrix with a center frequency of f2.

5. A method for improving the robustness of a sparse array of millimeter-wave radars according to claim 1, characterized in that, A set of data x obtained in Step 7 is expressed as: x = [ABS(x1)*exp(-1j*angle(x1),ABS(x2)*exp(-1j*angle(x2)) In the formula: ABS() is to take the absolute value, angle() is to take the phase, x1 is the data 2DFFT_1(rIdx,dIdx), which is the echo signal corresponding to the radar waveform with a center frequency of f1, and there is: where D1 = D, lmd1 is the wavelength of the radar waveform with a center frequency of f1, and θ is the angle; x2 is the data 2DFFT_2(rIdx,dIdx), which is the echo signal corresponding to the radar waveform with a center frequency of f2, and there is: where D2 = k×D1, lmd2 is the wavelength of the radar waveform with a center frequency of f2, and lmd2 = k×lmd1.

6. A method for improving the robustness of a sparse array of millimeter-wave radars as described in claim 5, characterized in that, After the said Step 7, it further includes: Step 8: Perform DBF angle measurement based on data x, and we have: θ e = argmax(A(θ)x) where θ e is the angle estimation value, and argmax(·) is the value of θ when the vector takes the maximum value, is the steering vector, and d ∈ [D1 ∪ D2].

Images