MIMO frequency modulation millimeter wave radar array reconfigurable array mode sharing method

Abstract

A reconfigurable array mode sharing method for an MIMO frequency modulation millimeter wave radar array comprises the steps that distance dimension fast Fourier transform and speed dimension fast Fourier transform are carried out on echo signal data matrixes obtained after analog-to-digital conversion is carried out on the basis of different modes, and distance Doppler spectrums of different receiving and transmitting antenna combinations are obtained; different modes with the same gain of the MIMO radar system are combined, the mode corresponding to the maximum virtual array aperture in the horizontal direction and the maximum virtual array aperture in the pitching direction under the gain is selected from the combination with the same gain, the two modes are subjected to angle solution and matching, angle solution of the maximum virtual array aperture is achieved in a combined mode, and the maximum virtual array aperture is obtained. A digital beam forming method is used for matching angle information of the two dimensions obtained by solving under a horizontal dimension virtual array and a pitching dimension virtual array with the maximum aperture. According to the invention, when a virtual array reconfigurable technology or an MIMO technology is used, the maximum array aperture in the original same gain mode can be effectively increased by using the complementary effect between different reconfigurable array modes.

Description

Technical Field

[0001] The present invention relates to a technology in the field of millimeter wave radar, and in particular to a reconfigurable array mode sharing method of a multiple-input multiple-output (MIMO) frequency-modulated continuous-wave millimeter wave radar L-shaped antenna array. Background Art

[0002] The existing MIMO millimeter-wave radar array reconfiguration technology mainly makes full use of the phase information between channels, that is, in the time division multiplexing (TDM) mode, multiple transmitting array elements or multiple receiving array elements are turned on at the same time, which effectively increases the diversity of virtual array configuration and forms a reconfigurable array, namely TR mn mode, m is the number of array elements that are simultaneously turned on in TX, and n is the number of array elements that are simultaneously turned on in RX. After being reconfigured, the system signal-to-noise ratio (SNR) and detection range can be effectively improved. However, in some modes, the reconfigurable array of the L-type array cannot maintain half the wavelength due to the array element spacing. The DOA super-resolution algorithm cannot accurately solve the target in two-dimensional space, which may lead to large angle errors in the azimuth or pitch dimensions, making it impossible to accurately locate the target in 4D space, and it is also impossible to form a target point cloud. Summary of the invention

[0003] The present invention aims to solve the problems that the existing MIMO technology has large or even wrong angle information solution of the target when the environmental noise causes a low signal-to-noise ratio (SNR), the MIMO radar has a limited detectable range of the target and can only detect targets at a shorter distance, and can only accurately resolve the angle in the pitch or azimuth dimension but not in the other dimension. A MIMO frequency-modulated millimeter-wave radar array reconfigurable array mode sharing method is proposed. While using virtual array reconfigurable technology or MIMO technology, the complementary effect between different reconfigurable array modes is utilized to effectively increase the maximum array aperture in the original same gain mode.

[0004] The present invention is achieved through the following technical solutions:

[0005] The present invention relates to a MIMO frequency modulation millimeter wave radar array reconfigurable array mode sharing method, comprising:

[0006] Step 1: In the initialization phase of the time-division multiplexing MIMO radar, different numbers of transmitting and receiving antennas are turned on at different times according to the needs, that is, the virtual arrays formed at different times are arranged differently, and the array of the traditional MIMO radar is reconstructed to achieve mode TR. mn , where m is the number of enabled transmit antennas and n is the number of enabled receive antennas.

[0007] Step 2: Perform range-dimensional fast Fourier transform and velocity-dimensional fast Fourier transform on the echo signal data matrix obtained after analog-to-digital conversion based on different modes, and obtain the range Doppler spectra of different transmitting and receiving antenna combinations.

[0008] Step 3, combining different modes with the same gain of the MIMO radar system, selecting the modes corresponding to the largest virtual array apertures in the horizontal and elevation directions under the same gain in the combination, respectively, performing angle resolution and matching on the two modes, and combining them to achieve the maximum virtual array aperture angle resolution, specifically: for the range Doppler spectrum of the mode corresponding to the largest virtual array aperture in the horizontal and elevation directions, a constant false alarm rate algorithm (CFAR) is used to obtain the distance and speed corresponding to the target, and a subspace method (TLS-ESPRIT) super-resolution wave angle estimation algorithm is used to solve the angle information of different target points to obtain the maximum virtual array aperture angle resolution.

[0009] Step 4: Use digital beamforming to match the angle information of the horizontal and elevation dimensions obtained by solving the maximum aperture virtual array in the two dimensions. Technical Effects

[0010] The present invention realizes the improvement of SNR and maximum detection range by turning on different numbers of transmitting antennas and receiving antennas to form MIMO radar arrays of different modes; by sharing reconfigurable array modes, the reconfigurable MIMO radar array technology is improved, the maximum aperture of the reconfigurable radar is expanded, and the angular resolution of the reconfigurable MIMO radar array is improved. Compared with the prior art, the present invention performs virtual array reconfiguration to improve the SNR and detection range of the MIMO architecture millimeter wave radar system, and can also solve the problem that some reconfigurable modes can only detect angle information in a single dimension of azimuth or pitch. After introducing the mode sharing method, the aperture of the largest array in the virtual reconfigurable array group with the same gain mode can be effectively increased, that is, the maximum aperture of the array of the single mode under the gain is further increased, and the spatial angle detection accuracy is improved. It can achieve the effect of increasing information of different dimensions of spatial angles, that is, obtaining accurate 2D spatial angle information and increasing the aperture of the largest virtual array in the virtual reconfigurable array group with the same gain mode under the virtual array reconfiguration condition. In the above-mentioned 4T8R L-type array, the maximum aperture of the reconfigurable array group with a normalized gain of 6 is increased from 4×5 to 7×7, and the maximum aperture of the reconfigurable array group with a normalized gain of 12 is increased from 3×4 to 5×6. The aperture of other gain groups can be increased in the same way. While improving the problem that some modes can only solve single-dimensional angle information, increasing the aperture effectively improves the angle estimation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic diagram of a reconfigurable array mode of an L-shaped array of the present invention;

[0012] Figure 2 It is a flow chart of the present invention;

[0013] Figure 3 is a block diagram of a MIMO FMCW radar system in an embodiment;

[0014] Figure 4 This is a schematic diagram of a 4×8L radar antenna array in an embodiment;

[0015] Figure 5 This is a schematic diagram of a 4×8 L-type radar antenna array MIMO virtual array in an embodiment;

[0016] Figure 6 This is a schematic diagram of a reconfigurable array of a 4×8 L-type radar antenna array in an embodiment;

[0017] Figure 7 It is a schematic diagram of the spatial wave model of the embodiment;

[0018] Figure 8 It is a schematic diagram of a reconfigurable array group mode sharing of a 4×8 L-type radar antenna array with a normalized gain of 6 in an embodiment;

[0019] Fig. 9 Schematic diagram of data sharing processing of a reconfigurable array group mode with normalized gain 6 in an embodiment;

[0020] Fig.10 A schematic diagram of a two-dimensional spatial angle coordinate system of an embodiment;

[0021] Fig.11 This is a graph showing the SNR-RMSE result of the two-dimensional spatial angle estimation of the reconfigurable array group with normalized gain 6 in the embodiment;

[0022] Fig.12 This is a graph of the SNR-RMSE results of two-dimensional spatial angle estimation of a reconfigurable array group with a normalized gain of 12 in the embodiment. Specific implementation methods

[0023] like Figure 3As shown, a MIMO frequency modulation millimeter wave radar array reconfigurable array mode sharing system for implementing the above method involved in this embodiment includes: a signal source 1 and a transmitting branch composed of a transmitting power divider 2, a power amplifier 3, and a transmitting antenna array TX respectively connected thereto, and a receiving branch composed of a receiving power divider 4, a low noise amplifier 5, and a receiving antenna array RX connected in sequence, wherein: the output end of the low noise amplifier 5 is sequentially provided with a low pass filter 6, an analog-to-digital converter 7 and a signal processor 8, the signal source 1 generates a transmitting signal, and a multi-channel transmitting signal is generated by the transmitting power divider 2 and passes through the power amplifier 3 to the port of each transmitting antenna TX; at the same time, the signal is scattered by the space target to generate an echo signal, which is received by the receiving antenna array RX and mixed with the signal of the receiving power divider 4, and the down-converted signal passes through the low noise amplifier 5 and the low pass filter 6, and is converted and sampled by the analog-to-digital converter 7, and then shared by the signal processor 8 in an L-type reconfigurable virtual array mode.

[0024] The signal processor 8 includes: a MIMO signal processing unit, an array reconfigurable unit, an array mode sharing unit, and a target space two-dimensional angle signal detection unit, wherein: the MIMO signal processing unit is connected to the radar ADC sampling data to process the most original radar signal, the array reconfigurable unit can reconfigure the MIMO architecture virtual array elements to form virtual arrays of different modes, the array mode sharing unit can form mode sharing for different and complementary reconfigurable array modes in the virtual reconfigurable array group of the same gain mode, and the target space two-dimensional angle information detection unit processes the received information and estimates the incoming wave angle.

[0025] like Figure 4 As shown, the transmitting antenna array TX and the receiving antenna array RX constitute a 4T8R L-type MIMO antenna array, wherein: the transmitting array is a three-unit uniform linear array with half-wavelength spacing in azimuth, and a two-unit uniform linear array with half-wavelength spacing in elevation; the receiving array element is a four-unit uniform linear array with 1.5 times wavelength spacing in azimuth, and a five-unit uniform linear array with one wavelength spacing in elevation.

[0026] like Figure 5 As shown, the L-shaped MIMO antenna array, through the virtual array equivalence principle, equates the MIMO antenna array of M transmitting antennas and N receiving antennas to a SIMO antenna array of 1 transmitting antenna M×N receiving antennas, that is, forms a MIMO radar virtual antenna array.

[0027] The 4T8R L-shaped array in this embodiment can equivalently form a uniform linear array with a maximum aperture of twelve elements and a half-wavelength in azimuth and a uniform linear array with ten elements and a half-wavelength in elevation, that is, the equivalent virtual array is a 10×12 L-shaped array, and a total of 21 passed data (one element is shared in azimuth and elevation) can be obtained by the digital-to-analog converter 7. The data received by the virtual array can be used to solve the pitch and azimuth angle information of the space target using the DOA algorithm.

[0028] like Figure 2 As shown, this embodiment involves a method for sharing reconfigurable array modes of a MIMO frequency-modulated millimeter-wave radar array of the above system, and the specific steps include:

[0001] Step 1: Collect the signal of the MIMO radar antenna array and generate a virtual array. After combining the data between different transmit and receive channels, a reconfigurable array is formed, namely TR mn mode, m is the number of TX array elements turned on at the same time, n is the number of RX array elements turned on at the same time, then the maximum sensing range of the system is System signal-to-noise ratio When multiple transmitting and receiving antennas are turned on, the overall array gain of the system By changing the MIMO virtual array working mode, the system perception range and signal-to-noise ratio (SNR) can be effectively improved. The modes with the same gain are defined as virtual reconfigurable array groups with the same gain mode, where: P t is the radar transmission power, σ is the RCS reflection cross-sectional area of ​​the target, G TX , G RX is the gain of the radar transmitting antenna and receiving antenna, λ is the radar operating wavelength, k is the Boltzmann constant, T0 is the operating temperature, B is the operating bandwidth, L is the loss factor, F n is the noise coefficient, and d is the distance between the target and the radar.

[0029] Step 2: The chirp signal emitted by the FMCW radar is S TX (t) = exp(j(2πf c t+πSt 2 +φ)), and the propagation speed of electromagnetic waves in space is the speed of light, which is much greater than the target's movement speed, so the target's movement distance that does not exceed the distance resolution during the signal transmission and reception process can be approximated, where f c is the signal carrier frequency, and the signal period is T c , bandwidth is B, then the modulation slope S = B / T c , φ is the initial phase of the signal, then the intermediate frequency signal obtained after receiving the signal and down-converting is Where: s Tx (t) and sRx (t) is the received and transmitted signal, A is the signal amplitude, R(τ) is the relative distance between the target and the radar, τ is the so-called "slow time", and c is the speed of light. The intermediate frequency signal is a single-frequency signal with a frequency of f IF =Sτ=2Sd / c, d is the distance between the target and the radar, that is, d=cf IF / 2S, the signal spectrum peak can be obtained by fast Fourier transform of the intermediate frequency signal, that is, the distance between the target and the radar. The phase difference between two adjacent chirps is Δφ≈2πf c t=4πvT c / λ, that is, the target's moving speed is: v = λΔφ / 4πT c , then the phase information at the peak frequency of the range spectrum can be fast Fourier transformed to obtain the phase change frequency and then solve the target's speed information. After receiving multiple transmitted chirp signals and down-converting them to obtain intermediate frequency signals, they are stored in a data matrix. The number of rows is the number of chirp signals, and the number of columns is the number of fast Fourier transformation points. By fast Fourier transforming each row, the peak frequency information can be obtained, and the distance information between the target and the radar can be solved. The data phase information at the peak frequency column is fast Fourier transformed to obtain the phase frequency, and the target's motion speed information can be solved. That is, based on the multi-target distance and speed information obtained, the multi-target range-doppler (RD) diagram can be obtained.

[0030] Step 3: Introduce the complementary characteristics between modes to reconstruct the shared array mode: combine the two modes of the horizontal maximum array aperture and the elevation maximum array aperture in the virtual reconfigurable array with the same gain mode, solve the angle information of the two dimensions and match them, specifically including: for each mode, obtain the distance and Doppler corresponding to the target through CFAR, such as Figure 7 As shown in the figure, the signal model is established. Since the angle information of the two dimensions is solved by a one-dimensional linear array, when there are M antenna array elements forming a one-dimensional linear array, the array element spacing is λ / 2, and there are K narrowband signal waves in the space, respectively [s1(t), s2(t), …, s K (t)], assuming that the spatial directions of these signals are [θ1(t), θ2(t), …, θ K (t)], when the first unit is the reference unit, the received signal of each unit is where n i (t) represents the additive noise of each antenna unit. The above model is represented by a matrix, that is, X = AS + N, where X(t) = [x1(t(, x2(t), ..., x M (t)] T is the array receiving signal, S(t) = [s1(t), s2(t), ..., sM (t)] T is the envelope of the signal, N(t) = [n1(t), n2(t), …, n M (t)] T is additive noise (stationary Gaussian random process), A = [a(θ1), a(θ2), a(θ K )] is the flow matrix of the signal, where is the phase difference vector. The TLS-ESPRIT algorithm is used to solve the one-dimensional angle for the virtual arrays with the largest aperture in two dimensions. The receiving signals of the first M-1 elements of the linear array are selected as the X sub-array, and the receiving signals of the last M-1 elements are selected as the Y sub-array. The received data can be expressed as: X(t) = AS(t) + n x (t), Y(t)=AΦS(t)+n y (t), where Φ is the phase difference matrix, and the two sub-matrices are combined to form Then use Z to obtain the covariance matrix Perform eigendecomposition on the covariance matrix and select the eigenvectors corresponding to the largest K eigenvalues ​​as the spatial signal subspace. Since the eigenvector signal subspace is the same as the space spanned by the array flow matrix, we have So there must exist a non-singular matrix T such that Established, so according to the construction of Z and the array flow matrix, we get Where E x represents the signal subspace obtained by the X array, E y represents the signal subspace obtained by the Y array, then E x Ψ=E y , considering the error in the actual data, that is So there is And it is necessary to make the correction matrix and vector R A , R B As small as possible, that is: Arranged It can be seen that [Ψ; -1] is The right singular vector corresponding to the minimum singular value. Therefore, under the action of correctly calibrating the matrix and vector, finding the correct right singular vector will obtain the required Ψ, that is, finding the matrix Make Minimum, while satisfying the constraints: FF H =I, as we know before, F is The right singular vector corresponding to the minimum singular value. According to the singular value solution method, Perform eigenvalue decomposition to find the eigenvectors corresponding to the smallest K eigenvalues, thereby forming a matrix F, which is divided into two upper and lower K×K matrices corresponding to F0 and F1. Combined with E x Ψ=E y Can get Where Ψ=T -1 ΦT, because T is a non-singular matrix, then Ψ and Φ are similar matrices, with common eigenvalues, and because Φ is a diagonal matrix, it can be said that Φ is the eigenvalue diagonal matrix of Ψ. Therefore, Φ can be obtained by performing eigenvalue decomposition on Ψ, and then the corresponding direction of arrival can be obtained through Φ, that is, the one-dimensional angle solution is completed.

[0031] Step 4: After step 3, the digital beamforming (DBF) method is used to match the angle information of the horizontal and elevation dimensions obtained under the maximum aperture virtual array.

[0032] The method of using digital beamforming is as follows: when the spacing of the equally spaced linear array is d, and the signal is incident at an angle θ to the normal of the array, the output signal of the array is Y=W T (θ0)X, where: X is the above received signal, W = [w0, w1, ..., w N ] is the signal steering vector, w i =2πd i sin(θ0 / λ), i=1, 2,...,N.

[0033] like Figure 8 As shown, the virtual reconfigurable array pattern of the 4×8L radar antenna array with normalized gain 6 is shared, and the normalized gain 6 array group includes TR 16 , TR 23 , TR 32 Three reconfigurable modes, system gain is MIMO radar TR 11 mode, and search for the maximum aperture array with half-wavelength spacing for the above three reconfigurable modes, TR 16 The mode has a four-element half-wavelength uniform linear array in elevation, a five-element half-wavelength uniform linear array in azimuth, and a TR 23 With only a 7-element half-wavelength uniform linear array in azimuth, it is impossible to accurately solve the target angle information in elevation. 32 The mode has a pitch seven-element half-wavelength uniform linear array, which cannot accurately solve the target angle information in azimuth. Only TR 16 The mode can accurately solve the two-dimensional spatial angle, while the other two modes cannot solve the two-dimensional angle information. The complementary method is used to introduce the common definition of the mode and dynamically configure the radar working mode. 23 The azimuth of the mode is 7-element half-wavelength uniform linear array and TR 32The pitch direction of the mode is composed of a seven-element half-wavelength uniform linear array to form an L-type array to solve the two-dimensional spatial angle information, that is, the mode shares virtual array elements, so that the two reconfigurable modes with complementary dimensions form a 7×7L-type virtual antenna array, which is more accurate than the TR in the gain group that can perform two-dimensional angle estimation. 11 The mode 4×5L virtual antenna array improves the maximum aperture and enables the two modes to have the ability to resolve the angle in two-dimensional space. The mode sharing method defined by the complementary principle can be applied between modes in any virtual reconfigurable array group with the same gain mode, which can form the largest aperture in the group and compensate for the resolution angle dimension.

[0034] like Fig. 9 As shown in the figure, the above-mentioned reconfigurable array group mode shared data processing process with normalized gain 6 is performed. By opening the positions of multiple transmitting and receiving array elements equivalent to virtual array elements, the elevation and azimuth of the two modes complement each other to form a 7×7L type half-wavelength uniform spacing array, TR 23 In the first chirp cycle, TX1 and TX4 are turned on at the same time to transmit signals. Channel 1 is formed by turning on RX1, RX5, and RX6 to receive data. Channel 2 is formed by turning on RX1, RX2, and RX7 to receive data. Channel 3 is formed by turning on RX1, RX4, and RX7 to receive data. Channel 4 is formed by turning on RX2, RX4, and RX7 to receive data. Channel 5 is formed by turning on RX3, RX4, and RX7 to receive data. Channel 6 is formed by turning on TX3 and TX4 to transmit signals at the same time in the second chirp cycle. Channel 7 is formed by turning on RX3, RX4, and RX7 to receive data. A reconfigurable seven-unit azimuth half-wavelength uniform virtual linear array is formed. The 7 passed data are stored in the data matrix. The number of rows is the number of channels, and the number of columns is the number of FFT points. TR 32 The same processing is performed on the two reconfigurable modes. Two data matrices are obtained by processing the data of the two reconfigurable modes respectively. The two-dimensional angle information is solved and matched using the super-resolution algorithm to obtain the two-dimensional angle information of the target space. That is, the performance of the 7×7L array under the normalized gain group can be achieved. Compared with the reconfigurable mode, the detection dimension of some modes is increased, the maximum array aperture in the gain group is increased, and the angle detection accuracy is improved.

[0035] like Fig.10 As shown in the figure, it is a schematic diagram of the two-dimensional angle selection of the spatial coordinate system. α and β in the coordinate system are selected as the spatial two-dimensional angle information of the target, and the two-dimensional spatial angle solving method based on TLS-ESPRIT is used to solve the angle information of the target.

[0036] like Fig.11 As shown in the figure, the RMSE of the two-dimensional spatial angle estimation of the reconfigurable array group with normalized gain 6 changes with SNR.23 and TR 32 It is impossible to solve the angle in both dimensions, so the error is very large. Then, for TR that can solve the angle 16 Mode and TR 23 -TR 32 The two modes share two arrays for simulation comparison, setting the spatial angle of the incoming wave to {18°, 24°}, {45°, 56°}, the radar bandwidth to 4 GHz, the number of snapshots to 256, the center frequency to 60 GHz, and assuming TR 11 The SNR is -10dB, and the SNR corresponding to the reconfigurable array group with normalized gain 6 is -2.2dB, so the simulation SNR is set from -2.2dB to 12.8dB, and 200,000 Monte Carlo simulations are performed at each SNR to calculate the RMSE of the estimated angle. It can be seen from the figure that after adopting mode sharing, the two modes in a virtual reconfigurable array group with the same gain mode can be used in a complementary manner and solved in different dimensions to obtain two-dimensional angle information, which not only solves the problem that some modes can only solve one-dimensional angle information, but also can obtain the maximum virtual array aperture of multi-mode combination in the virtual reconfigurable array group with the same gain mode, which increases the array aperture compared to single-mode operation.

[0037] like Fig.12 As shown in the figure, the RMSE of the two-dimensional spatial angle estimation of the reconfigurable array group with normalized gain 12 changes with the SNR. The virtual array formation and data processing flow of the virtual reconfigurable array group in this gain mode is the same as the above process. 34 and TR 43 It is impossible to solve the angle in both dimensions, so the error is very large. Then, for TR that can solve the angle 26 Mode and TR 34 -TR 43 The two modes share two arrays for simulation comparison, setting the spatial angle of the incoming wave to {18°, 24°}, {45°, 56°}, the radar bandwidth to 4 GHz, the number of snapshots to 256, the center frequency to 60 GHz, and assuming TR 11 The SNR is -10dB, and the SNR corresponding to the reconfigurable array group with normalized gain 12 is 0.8dB. The simulation SNR is set from 0.8dB to 15.8dB, and 200,000 Monte Carlo simulations are performed at each SNR to calculate the RMSE of the estimated angle. The conclusion is the same as above.

[0038] Compared with the prior art, the method utilizes the characteristics that different modes in the same gain mode virtual reconfigurable array group have the same gain but different virtual array apertures, introduces complementary characteristics, combines arrays of different modes, and forms the largest virtual array aperture in the gain virtual reconfigurable array group. Compared with the MIMO reconfigurable method, it can not only improve the radar's perception range and system SNR, but also improve the problem that some modes cannot be solved in two dimensions and increase the maximum virtual array aperture in the same gain mode virtual reconfigurable array group. For example, in the above 4T8R L-type array, the maximum aperture of the reconfigurable array group with normalized gain 6 is increased from 4×5 to 7×7, and the maximum aperture of the reconfigurable array group with normalized gain 12 is increased from 3×4 to 5×6. The apertures of other gain groups can also be increased in the same way, effectively improving the angle estimation accuracy.

[0039] In summary, the present invention fully utilizes the multi-channel data of the MIMO array for virtual array reconfiguration and at the same time introduces complementary characteristics, utilizes the characteristics that different modes in the virtual reconfigurable array group with the same gain mode have the same gain and different array apertures, performs mode sharing, combines the modes to form the largest aperture in the gain group, improves the array aperture compared to single-mode operation, and can improve the problem that some reconfigurable modes cannot solve two-dimensional angle information when working alone, thereby completing the pioneering mode sharing of the MIMO virtual reconfigurable array.

[0040] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. A method for sharing reconfigurable array modes of a MIMO frequency-modulated millimeter-wave radar array, characterized in that: include: Step 1: In the initialization phase of the time-division multiplexing MIMO radar, different numbers of transmitting and receiving antennas are turned on at different times according to the needs, that is, the virtual arrays formed at different times are arranged differently, and the array of the traditional MIMO radar is reconstructed to achieve mode TR. mn , where m is the number of enabled transmitting antennas and n is the number of enabled receiving antennas; Step 2: Performing distance-dimensional fast Fourier transform and speed-dimensional fast Fourier transform on the echo signal data matrix obtained after analog-to-digital conversion based on different modes, and obtaining the range Doppler spectrum of different transmitting and receiving antenna combinations; Step 3: Combine different modes with the same gain of the MIMO radar system, select the modes corresponding to the largest virtual array apertures in the horizontal and elevation directions under the same gain in the combination, respectively, resolve and match the two modes, and combine them to achieve the maximum virtual array aperture resolution; Step 4: Use digital beamforming to match the angle information of the horizontal and elevation dimensions obtained under the maximum aperture virtual array.

2. The MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method according to claim 1 is characterized by: The step 1 specifically includes: collecting signals from the MIMO radar antenna array and generating a virtual array, and forming a reconfigurable array by combining data between different transmitting and receiving channels, namely TR mn Mode, system maximum sensing range Signal-to-Noise Ratio When multiple transmitting and receiving antennas are turned on, the overall array gain of the system Where: m is the number of TX array elements turned on at the same time, n is the number of RX array elements turned on at the same time, P t is the radar transmission power, σ is the RCS reflection cross-sectional area of ​​the target, G TX ,G RX is the gain of the radar transmitting antenna and receiving antenna, λ is the radar operating wavelength, k is the Boltzmann constant, T0 is the operating temperature, B is the operating bandwidth, L is the loss factor, F n is the noise coefficient, and d is the distance between the target and the radar.

3. The MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method according to claim 1 is characterized by: The step 2 specifically includes: the chirp signal emitted by the FMCW radar is S TX (t) = exp(j(2πf c t+πSt 2 +φ)), the intermediate frequency signal is obtained after receiving the signal and down-converting it: The peak value of the signal spectrum is obtained by performing fast Fourier transform on the intermediate frequency signal, that is, the distance between the target and the radar is obtained. The phase information at the peak frequency of the distance spectrum is fast Fourier transformed to obtain the phase change frequency and then solve the target speed information. Multiple transmitted chirp signals are received and down-converted to obtain intermediate frequency signals and then stored in the data matrix. The number of rows is the number of chirp signals, and the number of columns is the number of fast Fourier transformation points. The peak frequency information can be obtained by performing fast Fourier transform on each row, and the distance information between the target and the radar can be solved. The data phase information at the peak frequency column is fast Fourier transformed to obtain the phase frequency, and the target movement speed information can be solved. That is, according to the multi-target distance and speed information obtained, the RD diagram of multiple targets is obtained, where: f c is the signal carrier frequency, and the signal period is T c , bandwidth is B, then the modulation slope S = B / T c ,φ is the initial phase of the signal, s Tx (t) and s Rx (t) is the received and transmitted signal, A is the signal amplitude, R(τ) is the relative distance between the target and the radar, τ is the slow time, c is the speed of light, the intermediate frequency signal is a single frequency signal, the frequency f IF =Sτ=2Sd / c, d is the distance between the target and the radar, that is, d=cf IF / 2S, the phase difference between two adjacent chirps is Δφ≈2πf c t=4πvT c / λ, that is, the target's moving speed is: v = λΔφ / 4πT c .

4. The MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method according to claim 1 is characterized by: The step 3 is specifically as follows: for the range Doppler spectrum of the mode corresponding to the largest virtual array aperture in the horizontal and pitch directions, the distance and speed corresponding to the target are obtained by a constant false alarm rate algorithm, and the angle information of different target points is solved by a subspace method super-resolution wave angle estimation algorithm to obtain the maximum virtual array aperture solution angle.

5. The MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method according to claim 1 or 4, characterized in that: For each mode, the distance and Doppler corresponding to the target are obtained through CFAR, and the signal model is established. Since the angle information of the two dimensions is solved using a one-dimensional linear array, when there are M antenna array elements forming a one-dimensional linear array, the array element spacing is λ / 2, and there are K narrowband signal waves in the space, respectively [s1(t), s2(t),…, s K (t)], assuming that the spatial directions of these signals are [θ1(t),θ2(t),…,θ K (t)], when the first unit is the reference unit, the received signal of each unit is The TLS-ESPRIT algorithm is used to solve the one-dimensional angle for the virtual arrays with the maximum aperture in two dimensions. The first M-1 array elements of the linear array are selected as the X subarray, and the last M-1 array elements are selected as the Y subarray. Then the received data X(t) = AS(t) + n x (t), Y(t)=AΦS(t)+n y (t), merge the two sub-arrays into Then use Z to obtain the covariance matrix Perform eigendecomposition on the covariance matrix and select the eigenvectors corresponding to the largest K eigenvalues ​​as the spatial signal subspace, where: i (t) represents the additive noise of each antenna unit. The above model can be represented by a matrix, that is, X = AS + N, X(t) = [x1(t), x2(t), …, x M (t)] T is the array receiving signal, S(t)=[s1(t),s2(t),…,s M (t)] T is the envelope of the signal, N(t)=[n1(t),n2(t),…,n M (t)] T is additive noise (stationary Gaussian random process), A=[a(θ1),a(θ2),,a(θ K )] is the flow matrix of the signal, is the phase difference vector, Φ is the phase difference matrix; The eigenvector signal subspace is the same as the space spanned by the array flow matrix, so we have So there must exist a non-singular matrix T such that Established, so according to the construction of Z and the array flow matrix, we get Where: E x The signal subspace obtained for the X array, E y is the signal subspace obtained by the Y array, then E x Ψ=E y , considering the error in the actual data, that is So there is And it is necessary to make the correction matrix and vector R A ,R B As small as possible, that is: Arranged Under the action of the correct calibration matrix and vector, finding the correct right singular vector will give the required Ψ, that is, finding the matrix Make Minimum, while satisfying the constraints: FF H =I, where: F is The right singular vector corresponding to the minimum singular value is calculated according to the singular value solution method. Perform eigenvalue decomposition and find the eigenvectors corresponding to the smallest K eigenvalues, thereby forming a matrix F, which is divided into two upper and lower K×K matrices corresponding to F0 and F1, thus obtaining Combined with E x Ψ=E y get Ψ=T -1 ΦT, where: Φ is the eigenvalue diagonal matrix of Ψ. Perform eigenvalue decomposition on Ψ to obtain Φ, and then find the corresponding direction of arrival for Φ, thus completing the one-dimensional angle solution.

6. The MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method according to claim 1 is characterized by: The step 4 specifically includes: in step 3, using the digital beamforming (DBF) method to solve and obtain corresponding angle information under the maximum aperture virtual array in the horizontal and elevation dimensions.

7. The MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method according to claim 6 is characterized by: The method of using digital beamforming is as follows: when the spacing of the equally spaced linear array is d, and the signal is incident at an angle θ to the normal of the array, the output signal of the array is Y=W T (θ0)X, where: X is the received signal, W=[w0,w1,…,w N ] is the signal steering vector, w i =2πd i sin(θ0 / λ),i=1,2,…,N.

8. A system for implementing the MIMO frequency modulated millimeter wave radar array reconfigurable array mode sharing method as described in any one of claims 1 to 7, characterized in that: include: A signal source and a transmitting branch composed of a transmitting power divider, a power amplifier, and a transmitting antenna array TX respectively connected thereto, and a receiving branch composed of a receiving power divider, a low noise amplifier, and a receiving antenna array RX connected thereto in sequence, wherein: the output end of the low noise amplifier is provided with a low pass filter, an analog-to-digital converter, and a signal processor in sequence, the signal source generates a transmitting signal, the transmitting power divider generates a multi-channel transmitting signal and passes through the power amplifier to the port of each transmitting antenna TX; at the same time, the signal is scattered by a space target to generate an echo signal, which is received by the receiving antenna array RX and mixed with the signal of the receiving power divider, and the down-converted signal passes through the low noise amplifier and the low pass filter, and is converted and sampled by the analog-to-digital converter and then shared by the signal processor in an L-shaped reconfigurable virtual array mode.

9. The system according to claim 8, characterized in that: The signal processor comprises: a MIMO signal processing unit, an array reconfigurable unit, an array mode sharing unit, and a target space two-dimensional angle signal detection unit, wherein: the MIMO signal processing unit is connected to the radar ADC sampling data to process the most original radar signal, the array reconfigurable unit can reconfigure the MIMO architecture virtual array elements to form virtual arrays of different modes, the array mode sharing unit can form mode sharing for different and complementary reconfigurable array modes in the virtual reconfigurable array group of the same gain mode, and the target space two-dimensional angle information detection unit processes the received information and estimates the incoming wave angle.

10. The system according to claim 8, characterized in that: The transmitting antenna array TX and the receiving antenna array RX constitute a 4T8R L-type MIMO antenna array, wherein: the transmitting array is a three-unit uniform linear array with half-wavelength spacing in azimuth, and a two-unit uniform linear array with half-wavelength spacing in elevation; the receiving array element is a four-unit uniform linear array with 1.5 times wavelength spacing in azimuth, and a five-unit uniform linear array with one wavelength spacing in elevation; The L-shaped MIMO antenna array uses the virtual array equivalence principle to convert a MIMO antenna array of M transmitting antennas and N receiving antennas into a SIMO antenna array of 1 transmitting antenna and M×N receiving antennas, thus forming a MIMO radar virtual antenna array.

Images