High-precision angle estimation method for distributed millimeter wave radar network, medium and equipment

Abstract

The invention provides a high-precision angle estimation method for a distributed millimeter wave radar network, a medium and equipment, and belongs to the technical field of radar signal processing. According to the method, a large-aperture MIMO radar system is constructed based on a plurality of small millimeter wave radars, and an initial rough angle value of a target is obtained by using two-dimensional fast Fourier transform; accurately calculating the walking position of the distance spectrum peak value of each virtual antenna according to the distance resolution of the radar and the rough angle value, extracting all distance spectrum peak value data according to the walking position, and generating a one-dimensional angle vector; performing angle dimension fast Fourier transform on the one-dimensional angle vector to generate an angle spectrum; calculating a target angle estimation value according to the angle spectrum peak value position; the above process is iteratively circulated until the error between two adjacent angle estimation values is smaller than a preset threshold value. According to the method, different types of distributed millimeter wave radar network architectures can be flexibly adapted, the precision problem caused by target distance walking is effectively eliminated, and the angle estimation precision is greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of radar signal processing, specifically relating to a high-precision angle estimation method, medium, and device for distributed millimeter-wave radar networks. Background Technology

[0002] As radar detection systems evolve towards networking and intelligence, distributed millimeter-wave radar networks, with their high angular resolution, multi-node collaborative sensing capabilities, and strong anti-interference characteristics brought by large-aperture virtual arrays, have become the core development direction of next-generation radar systems. As a key technological support, multiple-input multiple-output (MIMO) radar significantly improves target positioning accuracy through virtual aperture synthesis mechanisms. However, angle estimation faces severe challenges in distributed millimeter-wave radar networks: on the one hand, traditional angle estimation methods (such as digital beamforming based on the plane wave assumption) neglect the spherical wavefront curvature effect, leading to range migration and angle errors exceeding 4° at 500m; on the other hand, the geometric discontinuity of the virtual array broadens the Fast Fourier Transform (FFT) spectrum, resulting in a computational complexity of O(N) for subspace algorithms (such as MUSIC). 3 This makes it difficult to meet real-time processing requirements. In addition, existing methods mostly rely on preset array topologies (such as uniform linear arrays), which cannot adapt to topology changes in dynamic deployment scenarios such as vehicles and drones, severely restricting the realization of the potential of the millimeter-wave band.

[0003] Against this backdrop, there is an urgent need to develop a new angle estimation method that combines adaptive distance movement compensation, low-complexity processing, and topology independence, in order to overcome the bottleneck of phase mismatch caused by path difference in distributed systems and provide core technical support for high-precision positioning. Summary of the Invention

[0004] This invention addresses the problems of large distance travel error, spectral broadening caused by large virtual aperture, and phase mismatch caused by network topology changes in traditional angle estimation methods for distributed millimeter-wave radar networks. It provides a high-precision angle estimation method, medium, and device for distributed millimeter-wave radar networks. While maintaining the performance advantages of the millimeter-wave band, the computational complexity is significantly reduced, adapting to the real-time processing requirements of heterogeneous hardware units, and providing an efficient solution for high-precision, unambiguous positioning of long-range multi-targets.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a high-precision angle estimation method for distributed millimeter-wave radar networks, comprising the following steps:

[0007] S1: Based on a MIMO radar system formed by multiple millimeter-wave radars, a large-aperture virtual antenna array is constructed, and the initial roughness angle value of the target is obtained by using two-dimensional fast Fourier transform on the echo signal.

[0008] S2: Based on the initial coarse angle value in step S1 and the radar range resolution, calculate the range spectrum peak offset of each virtual antenna, and extract all range spectrum peak data accordingly to generate a one-dimensional angle vector.

[0009] S3: Perform an angle-dimensional fast Fourier transform on the one-dimensional angle vector from step S2 to generate an angle spectrum, and calculate the target angle estimate based on the peak position of the angle spectrum;

[0010] S4: Use the target angle estimate as the new initial rough angle value, return to step S2 for iterative loop until the error between two adjacent target angle estimates is less than the preset threshold or the maximum number of iterations is reached, and output the final target angle estimate.

[0011] Optionally, in step S1, the construction of a large-aperture virtual antenna array based on the MIMO radar system formed by multiple millimeter-wave radars specifically involves:

[0012] Construct a distributed MIMO radar system consisting of multiple millimeter-wave radars, with each radar node deployed in an arbitrary transmit / receive geometric topology to form a large-aperture virtual antenna array.

[0013] Optionally, in step S1, obtaining the initial roughness angle value of the target using a two-dimensional fast Fourier transform on the echo data specifically involves:

[0014] The frequency-modulated continuous wave echo signals of all radar channels of the large-aperture virtual antenna array are collected. A two-dimensional fast Fourier transform is performed on the echo signal of each channel to generate a range spectrum-channel matrix X(f, v), where f represents the range spectrum and v represents the virtual channel index.

[0015] The distance-averaged spectral vector is obtained by averaging along the channel dimension. The target distance peak frequency f is obtained through peak detection. r ;

[0016] Extracting the distance spectrum from the distance spectrum-channel matrix X(f, v) as f r A one-dimensional angle vector X r (v) Perform a Fast Fourier Transform in the angle dimension to generate an angle spectrum, and obtain the target angle peak frequency f through peak detection. a The initial roughness angle value θ of the target is determined as follows:

[0017]

[0018] In the formula, c represents the speed of light, f0 represents the center frequency of the transmitted signal, and d represents the spacing between the virtual array antennas.

[0019] Optionally, step S2 is as follows:

[0020] Based on the initial roughness angle value θ and the radar range resolution d res Calculate the peak offset ε of the range spectrum for each virtual antenna. index (v):

[0021]

[0022] The peak frequency f of the range spectrum of the virtual antenna v is calculated based on the following compensation model. r (v):

[0023] f r (v)=f r +ε index (v);

[0024] Extract f sequentially r (v) The channel data is used as the distance spectrum peak data. The distance spectrum peak data is sorted according to spatial location to generate a one-dimensional angle vector.

[0025] Optionally, step S3 is as follows:

[0026] Perform a Fast Fourier Transform (FFT) on a one-dimensional angle vector, expanding the number of transform points to 1024, to locate the peak position f of the angle spectrum. a The estimated target angle θ′ is calculated as follows:

[0027]

[0028] Optionally, in step S4, the estimated target angle θ for each iteration is recorded. k Calculate the absolute error ε between two consecutive target angle estimates:

[0029] ε=|θ k -θ k-1 |;

[0030] In the formula, θ k-1 This represents the target angle estimate from the previous iteration;

[0031] When the error ε is less than the preset threshold or the maximum number of iterations is reached, the iteration is terminated, and the final target angle estimate is output.

[0032] In a second aspect, the present invention provides a computer-readable storage medium storing a computer program that causes a computer to execute the high-precision angle estimation method for distributed millimeter-wave radar networks as described in the first aspect.

[0033] Thirdly, the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the high-precision angle estimation method for distributed millimeter-wave radar networks as described in the first aspect.

[0034] The beneficial effects of this invention are:

[0035] (1) This invention achieves high-precision estimation of target angle in a sparsely deployed distributed millimeter-wave radar network by using a large-aperture virtual array and iterative range travel compensation. Specifically, it brings three key gains: eliminating spectral peak drift caused by path difference; significantly improving angle resolution and estimation accuracy; and converging to a high-confidence angle solution while maintaining low complexity.

[0036] (2) The present invention adopts a closed-loop iterative optimization architecture. Through the adaptive progressive mechanism of coarse-fine estimation, it supports the collaborative processing of heterogeneous radar units (such as different position accuracy or clock synchronization levels) and adapts to the dynamic reconstruction of network topology (such as vehicle radar array deformation), thus overcoming the limitations of traditional methods that rely on preset geometry. Attached Figure Description

[0037] Figure 1 A flowchart of a high-precision angle estimation method for distributed millimeter-wave radar networks;

[0038] Figure 2 A schematic diagram of the radar range spectrum-channel matrix for a target distance of 5m and an azimuth angle of -70°.

[0039] Figure 3 The angle spectrum and angle estimates are obtained using the traditional 2DFFT method;

[0040] Figure 4 The angle spectrum and angle estimates obtained using the method of this invention;

[0041] Figure 5 The absolute angle estimation error curves are shown for the traditional method and the method of this invention, with a radar range resolution of 0.0593m.

[0042] Figure 6 The absolute angle estimation error curves are shown for the traditional method and the method of this invention, with a radar range resolution of 0.0297m. Detailed Implementation

[0043] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0044] In one embodiment, the present invention proposes a high-precision angle estimation method for distributed millimeter-wave radar networks, such as... Figure 1 As shown, the method includes the following steps:

[0045] S1: Based on a MIMO radar system composed of multiple millimeter-wave radars, a large-aperture virtual antenna array is constructed, and the target's range and initial roughness angle values ​​are obtained from the echo data using a two-dimensional fast Fourier transform.

[0046] In this embodiment, firstly, a distributed MIMO radar system consisting of multiple small millimeter-wave radars is constructed, with each radar node deployed in an arbitrary transmit / receive geometric topology to form a large-aperture virtual array.

[0047] Then, frequency-modulated continuous wave echo signals from all radar channels are acquired, and a range-dimensional fast Fourier transform is performed on the data of each channel to generate a range spectrum-channel matrix X(f, v), where f represents the range spectrum and v represents the virtual channel index.

[0048] Next, the distance-averaged spectral vector is obtained by averaging along the channel dimension. The target distance peak frequency f is obtained through peak detection. r Determine the distance to the target:

[0049]

[0050] Where c represents the speed of light, and α represents the frequency modulation slope of the emitted signal;

[0051] Finally, the distance spectrum f is extracted from the distance spectrum-channel matrix X(f, v). r A one-dimensional angle vector X r (v) is subjected to an angle-dimensional Fast Fourier Transform to generate an angle spectrum. The target angle peak frequency f is obtained through peak detection. a Determine the initial roughness angle value θ of the target:

[0052]

[0053] Where f0 represents the center frequency of the transmitted signal, and d represents the spacing between the virtual array antennas.

[0054] S2: Based on the initial coarse angle value from step S1 and the radar's range resolution, accurately calculate the movement position of the range spectrum peak of each virtual antenna, and extract all range spectrum peak data accordingly to generate a one-dimensional angle vector.

[0055] In this embodiment, based on the initial roughness angle value θ obtained in step S1 and the radar range resolution d res Calculate the range spectrum peak offset ε caused by the path difference for each virtual antenna. index ,as follows;

[0056]

[0057] Recalculate the target range peak frequency f for each virtual antenna v. r (v), the specific compensation model is as follows:

[0058] f r (v)=f r +ε idex (v)

[0059] Then extract the distance spectrum f sequentially. r (v) The channel data it is located in.

[0060] S3: Perform an angle-dimensional Fast Fourier Transform (Angle-FFT) on the one-dimensional angle vector to generate an angle spectrum. Finally, calculate the target angle estimate based on the peak positions of the angle spectrum.

[0061] In this embodiment, all virtual antenna compensation data extracted in S2 are sorted by spatial location to construct a one-dimensional angle vector. A high-resolution angle-dimensional fast Fourier transform is then performed on this vector, expanding the number of transform points to 1024 to improve spectral resolution and locate the spectral peak position f. a Estimate the high-precision angle θ′:

[0062]

[0063] S4: Use the angle value output in step S3 as the new initial rough angle value, and iterate through steps S2-S3 until the error between two adjacent angle estimation results is less than the preset threshold or the maximum number of iterations is reached. Finally, output a high-precision angle estimate.

[0064] In this embodiment, the angle estimate θ′ output in step S3 is used as the new initial rough angle value and input into step S2 to re-execute the distance travel compensation and angle estimation process. The angle estimation result θ is recorded in each iteration. k Calculate the absolute error ε between two consecutive estimates:

[0065] ε=|θ k -θ k-1 |

[0066] When the error ε is less than the preset threshold (typically the angle resolution d), θ The iteration terminates when the maximum number of iterations (usually 5) is reached. The final output is a high-precision angle estimate.

[0067] Next, we will use specific experiments to illustrate the effectiveness of the method proposed in this invention.

[0068] Figure 2The radar range spectrum-channel matrix is ​​shown in a scenario with a target distance of 5m and an azimuth angle of -70°. Under the conditions of radar range resolution of 0.0593m and virtual antenna aperture of 80 [half wavelength], the range spectrum peak shift phenomenon caused by the large aperture effect can be clearly observed—the target peak position shifts regularly with the virtual antenna channel index (the maximum shift reaches 4.2 range cells), which intuitively verifies the phase mismatch problem caused by range movement.

[0069] Figure 3 and Figure 4 The angle spectrum and angle estimation results obtained using the traditional 2DFFT method and the method of this invention are presented respectively. Figure 3 The rough angle estimate is -72.76° (absolute error 2.76°), and its angle spectrum peak intensity is 1.367 × 10⁻⁶. 4 The method of this invention ( Figure 4 The precise estimate of the spectral peak intensity is -69.96° (absolute error 0.04°), and the peak intensity is increased to 1.548 × 10⁻⁶. 4 (Gain 13%) indicates that the distance movement compensation mechanism significantly improves the array response signal-to-noise ratio and phase consistency.

[0070] Figure 5 The results show the absolute angle estimation error at a radar range resolution of 0.0593m. It can be seen that as the target angle changes from 0° to -90° (90°), the error of the coarse angle estimate obtained by the traditional 2DFFT method increases, even exceeding 4°, while the error of the method of this invention is less than 0.5° across the entire angle, improving accuracy and stability by 8 times.

[0071] Figure 6 The results show the absolute angle estimation error at a radar range resolution of 0.0297m. The trend of the angle roughness error obtained using the traditional 2DFFT method is also presented. Figure 5 Similarly, but the maximum error increased to 6° (a 50% deterioration compared to conventional resolution), proving that increasing resolution exacerbates the distance-travel effect. However, the method of this invention maintains an angle estimation accuracy error of no more than 0.5° even with increased resolution, breaking through the bottleneck of the resolution-accuracy trade-off.

[0072] In summary, this invention overcomes the distance movement problem in distributed millimeter-wave radar networks by using a dynamic path difference compensation model and a closed-loop iterative optimization architecture. It not only improves the accuracy of angle estimation but also enhances the angle spectrum signal-to-noise ratio, providing a reliable technical solution for real-time high-precision positioning of multiple targets in complex electromagnetic environments. It is particularly suitable for dynamic topology scenarios such as intelligent driving and drone formations.

[0073] In another embodiment, the present invention provides a computer-readable storage medium storing a computer program that causes a computer to execute the high-precision angle estimation method for distributed millimeter-wave radar networks described in the foregoing embodiments.

[0074] In another embodiment, the present invention proposes an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the high-precision angle estimation method for distributed millimeter-wave radar networks described in the foregoing embodiments.

[0075] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0076] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0077] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A high-precision angle estimation method for distributed millimeter-wave radar networks, characterized in that, Includes the following steps: S1: Based on a MIMO radar system formed by multiple millimeter-wave radars, a large-aperture virtual antenna array is constructed, and the initial roughness angle value of the target is obtained by using two-dimensional fast Fourier transform on the echo signal. S2: Based on the initial coarse angle value in step S1 and the radar range resolution, calculate the range spectrum peak offset of each virtual antenna, and extract all range spectrum peak data accordingly to generate a one-dimensional angle vector. S3: Perform an angle-dimensional fast Fourier transform on the one-dimensional angle vector from step S2 to generate an angle spectrum, and calculate the target angle estimate based on the peak position of the angle spectrum; S4: Use the target angle estimate as the new initial rough angle value, return to step S2 for iterative loop until the error between two adjacent target angle estimates is less than the preset threshold or the maximum number of iterations is reached, and output the final target angle estimate.

2. The high-precision angle estimation method for distributed millimeter-wave radar networks as described in claim 1, characterized in that: In step S1, the construction of a large-aperture virtual antenna array based on a MIMO radar system formed by multiple millimeter-wave radars specifically involves: Construct a distributed MIMO radar system consisting of multiple millimeter-wave radars, with each radar node deployed in an arbitrary transmit / receive geometric topology to form a large-aperture virtual antenna array.

3. The high-precision angle estimation method for distributed millimeter-wave radar networks as described in claim 2, characterized in that: In step S1, obtaining the initial roughness angle value of the target using a two-dimensional fast Fourier transform on the echo data specifically involves: The frequency-modulated continuous wave echo signals of all radar channels of the large-aperture virtual antenna array are collected. A two-dimensional fast Fourier transform is performed on the echo signal of each channel to generate a range spectrum-channel matrix X(f, v), where f represents the range spectrum and v represents the virtual channel index. The distance-averaged spectral vector is obtained by averaging along the channel dimension. The peak frequency f of the target distance is obtained through peak detection. r ; Extracting the distance spectrum from the distance spectrum-channel matrix X(f, v) as f r A one-dimensional angle vector X r (v) Perform a Fast Fourier Transform in the angle dimension to generate an angle spectrum, and obtain the target angle peak frequency f through peak detection. a The initial roughness angle value θ of the target is determined as follows: In the formula, c represents the speed of light, f0 represents the center frequency of the transmitted signal, and d represents the spacing between the virtual array antennas.

4. The high-precision angle estimation method for distributed millimeter-wave radar networks as described in claim 3, characterized in that: Step S2 is as follows: Based on the initial roughness angle value θ and the radar range resolution d res Calculate the peak offset ε of the range spectrum for each virtual antenna. index (v): The peak frequency f of the range spectrum of the virtual antenna v is calculated based on the following compensation model. r (v): f r (v)=f r +ε index (v); Extract f sequentially r (v) The channel data is used as the distance spectrum peak data. The distance spectrum peak data is sorted according to spatial location to generate a one-dimensional angle vector.

5. The high-precision angle estimation method for distributed millimeter-wave radar networks as described in claim 4, characterized in that: Step S3 is as follows: Perform a Fast Fourier Transform (FFT) on a one-dimensional angle vector, expanding the number of transform points to 1024, to locate the peak position f of the angle spectrum. a The estimated target angle θ′ is calculated as follows:

6. The high-precision angle estimation method for distributed millimeter-wave radar networks as described in claim 1, characterized in that: In step S4, the estimated target angle θ is recorded for each iteration. k Calculate the absolute error ε between two consecutive target angle estimates: ε=|θ k -θ k-1 |; In the formula, θ k-1 This represents the target angle estimate from the previous iteration; When the error ε is less than the preset threshold or the maximum number of iterations is reached, the iteration is terminated, and the final target angle estimate is output.

7. A computer-readable storage medium storing a computer program, characterized in that, The computer program causes the computer to execute the high-precision angle estimation method for distributed millimeter-wave radar networks as described in any one of claims 1-6.

8. An electronic device, characterized in that, include: The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the high-precision angle estimation method for distributed millimeter-wave radar networks as described in any one of claims 1-6.

Images