Coherent signal source single-snapshot super-resolution angle estimation method and system based on Toeplitz matrix

Abstract

The invention discloses a coherent signal source single-snapshot super-resolution angle estimation method and system based on a Toeplitz matrix, and the method comprises the steps: carrying out pulse compression and Doppler processing, obtaining the distance information and speed information of a target, extracting single-snapshot channel data, and calculating a covariance matrix; a Toeplitz matrix is used to reconstruct the covariance matrix, and decoherence processing is completed; and super-resolution angle estimation is carried out through a beam domain MUSIC algorithm. The method aims to solve the problem that a strong target and a weak target which are the same in speed, the same in distance but different in angle coincide on an RD spectrum, the limitation of a traditional method on the angle resolution is broken through, accurate angle estimation of a coherent signal source is achieved, and the method is particularly suitable for the situation that DOA estimation cannot be conducted through digital beam forming and conventional MUSIC. According to the invention, when the channel signal-to-noise ratio is-20dB, the angle estimation error is less than 1 degree, and the precision and robustness of the system are significantly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of radar direction of arrival (DOA) estimation, and in particular to a Toeplitz matrix-based coherent source single snapshot super-resolution angle estimation (TB-MUSIC) method and system. Background Art

[0002] In radar array signal processing, increasing the number of array elements generally improves the resolution of angle estimation. When the number of array elements is limited, super-resolution angle estimation methods are needed to improve angular resolution. For traditional radar array signal processing methods, the mainlobe width is often affected by the number of array elements. With a certain number of array elements, the mainlobe width is limited, making it impossible to accurately resolve multiple signal sources with similar angles. If the angular differences between signal sources are small, the limited mainlobe width will prevent these sources from being effectively resolved. Super-resolution angle estimation methods can improve angular resolution with a lower number of array elements and a wider mainlobe. In practical applications, super-resolution angle estimation can significantly improve radar system performance. Even when the number of array elements is limited, super-resolution algorithms can mitigate the performance degradation caused by insufficient array elements. They can provide better signal source localization and direction estimation than traditional methods, enabling these systems to more accurately acquire target information.

[0003] Single-snapshot data is commonly used in scenarios with high real-time requirements. It can be quickly processed and output, reducing latency. Since only a single snapshot's echo signal needs to be processed, the computational burden is relatively light. This meets the system's real-time requirements and effectively reduces the system's computational complexity.

[0004] In the case of multiple targets, coherent or nearly coherent echo signals may be generated between targets. Without signal decorrelation, target signals will experience aliasing in angle estimation, making it impossible for the radar system to distinguish between closely spaced targets. Decorrelation separates the echo signals from different targets, improving the system's ability to distinguish multiple close targets and thus enhancing the accuracy of Direction of Access (DOA) estimation. Decorrelation enhances the robustness of signal processing algorithms, especially in situations with significant target overlap or multipath propagation. By decorrelation, signal interference caused by multipath propagation and reflections can be effectively reduced, making DOA estimation more reliable. This is particularly important for radar systems operating in complex environments (such as urban, mountainous, and ocean environments).

[0005] If the signal of one target is much stronger than that of another, the presence of the strong signal may completely drown out or blur the signal of the weaker target, making it impossible to correctly separate the two in angle estimation. This can lead to errors in angle estimation, especially when the targets are coherent. The presence of a strong signal target may affect the width and shape of the array's main lobe, causing the echo signal of the weaker signal target to be suppressed or unable to be correctly detected. Especially when utilizing traditional beamforming technology, the directivity of the array may become less than ideal due to the influence of the strong signal, thus affecting the angle estimation of the weaker signal target. In addition, if the signal of one target is very strong and the signal of the other target is very weak, this difference in signal strength will cause the signal-to-noise ratio of the weaker signal target to drop significantly, and the angle estimation error of the weaker signal target will increase significantly. The strong signal target may completely dominate the signal processing, affecting the overall performance and accuracy of the algorithm. Therefore, when solving the problem of coherent target angle estimation, it is of great significance to achieve accurate angle estimation for two coherent targets, one strong and one weak. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this paper provides a single-snapshot super-resolution angle estimation method and system for coherent signal sources based on the Toeplitz matrix. This method employs the TB-MUSIC algorithm to achieve angle estimation for coherent signal sources, addressing the phenomenon in which two strong and weak targets with the same speed, distance, and angle overlap on the radar spectrum. To address the problem of coherent signal sources being unable to be estimated using digital beamforming and conventional MUSIC, the TB-MUSIC algorithm is employed to address this problem.

[0007] In order to achieve the above object of the invention, the technical solution adopted by the present invention is as follows:

[0008] A method for super-resolution angle estimation of a coherent source single snapshot based on a Toeplitz matrix comprises the following steps:

[0009] S1: Perform pulse compression and Doppler processing on the radar echo signal of a complete coherent integration cycle to obtain the target's range and velocity information, and obtain the Channel Range Doppler Spectrum (CRD) spectrum. Fix the target's range unit and Doppler unit, extract the single-snapshot channel data from the CRD spectrum, and calculate the single-snapshot covariance matrix.

[0010] S2: Based on the obtained single-snapshot covariance matrix, the Toeplitz matrix is used to reconstruct the covariance matrix of the single-snapshot channel data to achieve decorrelation processing;

[0011] S3: The covariance matrix of the decorrelated single snapshot channel data is used as input, and the angle of the coherent source is estimated by the beam-domain MUSIC algorithm.

[0012] Furthermore, the pulse compression process processes the pulse signal through a matched filter to determine the target's distance information through the peak value; and the Doppler processing obtains the range-Doppler-channel spectrum through Fourier transform to clarify the target's distance and speed information.

[0013] Furthermore, in S1, assuming that the radar's transmitted signal is a linear frequency modulation signal, the transmitted signal during the entire coherent accumulation period is:

[0014]

[0015] Where H is the number of pulses, T r is the pulse repetition period, f c is the carrier frequency, k = B / T is the frequency modulation slope, B is the bandwidth, and T is the pulse width. The number of points in each pulse is M = f s ·T r , f s is the sampling frequency, T / T r The duty cycle is set to 1 / 15.

[0016] The radar receiving signal is:

[0017]

[0018] Where τ is the target delay, which is related to the distance R and speed v:

[0019]

[0020] The antenna array uses a uniform linear array of N elements. Taking the leftmost antenna as the reference, the N×1 dimensional spatial steering matrix is:

[0021] a(θ)=[0,...,e j2π(N-1)dsinθ / λ ] T

[0022] Where λ is the wavelength, d = λ / 2 is the array element spacing, and θ is the target location.

[0023] Assume there are two coherent targets in space, with the same distance and speed, but different directions, and one with a strong amplitude and the other with a weak amplitude. The received signal is:

[0024] S r (t)=a(θ0)·S0·S r0 (t)+a(θ1)·S1·S r0 (t)+n(t)

[0025] Where S0 and S1 are the amplitudes of the two targets, S0>S1, θ0 is the angle of target 1, θ1 is the angle of target 2, and n(t) is Gaussian white noise.

[0026] Furthermore, the range Doppler processing in S1 includes:

[0027] The radar signal of the entire coherent accumulation period is transformed into three dimensions with the dimensions H·M·N, where H is the number of pulses, M is the number of points of a single pulse, and N is the number of array elements.

[0028] The basic idea of distance processing is to perform H pulse compression calculations on H fast time-dimensional pulses through pulse compression, and determine the distance information through the peak value. Pulse compression is to set a matched filter, which is the deconvolution and conjugation of the transmitted signal, specifically:

[0029] y R (t) = S r (t)*h(t)

[0030] h(t)=S o * (-t)

[0031] Where * represents convolution. As can be seen from the above formula, the phase-frequency characteristics of the matched filter are opposite to those of the received signal. Through convolution, the signal's phase can be reduced to zero, achieving coherent accumulation. Conversely, the phase of noise is random, so only non-coherent accumulation can be achieved, and the peak value can be used to clearly determine the target's distance information.

[0032] The pulse compression process is implemented using the frequency domain method:

[0033] S r (w) = fft[S r (t)],H(w)=fft[h(t)]

[0034] Y R (w)=S r (w)·H(w),y R (t) = ifft[Y R (w)]

[0035] The received signal and the matched filter can be transformed into the frequency domain through Fourier transform, multiplied and then transformed back into the time domain using inverse Fourier transform to complete the distance processing.

[0036] Doppler processing is performed on the radar echo signal in S1:

[0037] Obtain range-Doppler-channel spectrum to clarify distance and velocity information.

[0038] y RD (t) = fft[y R (t)·w hamming ] / H

[0039] Among them, w hamming is a Hamming window of length H.

[0040] Extract N×1-dimensional single-snapshot channel data from the RD spectrum and calculate its covariance matrix:

[0041] R=y RDC ·y RDC H

[0042] Among them, y RDC is a single snapshot channel data, R is the N×N dimensional covariance matrix, [·] H represents the conjugate transpose.

[0043] Furthermore, in S2, the Toeplitz matrix is used to perform decorrelation processing on the covariance matrix of the single snapshot channel data. The specific method is:

[0044] The Toeplitz matrix is constructed using the covariance matrix of the single snapshot channel data, so that the newly constructed covariance matrix is full rank.

[0045] The structure of the Toeplitz matrix is:

[0046]

[0047] Among them, a0~a n-1 is the first column of the covariance matrix of the single snapshot channel data, a0~a -(n-1) It is the first row of the covariance matrix of the single snapshot channel data.

[0048] Furthermore, in S3, the covariance matrix obtained after the decorrelation processing is used to perform the beam domain MUSIC algorithm to perform super-resolution angle estimation. The specific method is as follows:

[0049] Convert channel information into beam information and transform data into beam domain using conversion matrix. The dimension of the conversion matrix is M×N. B , which has the form:

[0050]

[0051] Among them, N B is the number of beams, is the angle of the composite beam.

[0052] Using the covariance matrix and transformation matrix reconstructed by the Toeplitz matrix, the data is transformed into the beam domain to obtain the covariance matrix in the beam domain:

[0053]

[0054] Among them, R B is the covariance matrix in the beam domain, whose dimension is N B ×N B .

[0055] Using the covariance matrix in the beam domain, the MUSIC method is used to estimate the angle and perform eigenvalue decomposition to obtain eigenvalues and eigenvectors. The eigenvectors are then rearranged in descending order of eigenvalues. The formula is as follows:

[0056] R B =UΣU H

[0057] sort(U)sort(Σ)

[0058] Among them, Σ is the eigenvalue, U is the eigenvector, and it is divided into signal subspace and noise subspace according to the number of sources. The first S columns with larger eigenvalues represent the signal subspace, and the last N columns with smaller eigenvalues represent the noise subspace. B - The S column represents the noise subspace.

[0059] Generate a spatial compensation vector:

[0060]

[0061] The noise subspace, transformation matrix and spatial compensation vector are used to find the spatial spectrum estimation coefficients:

[0062]

[0063] Where θ is the azimuth angle of the target, n is the number of array elements, d is the element spacing, and λ is the wavelength.

[0064] The present invention also discloses a coherent source single snapshot super-resolution angle estimation system based on a Toeplitz matrix. The system can be used to implement the above-mentioned coherent source single snapshot super-resolution angle estimation method, specifically comprising:

[0065] Pulse compression and Doppler processing module: used to perform pulse compression and Doppler processing on the radar echo signal of a complete coherent integration cycle, obtain the target's range and velocity information, obtain the CRD spectrum, fix the target's range unit and Doppler unit, extract the single-snapshot channel data from the CRD spectrum, and calculate the single-snapshot covariance matrix;

[0066] Toeplitz matrix reconstruction module: Based on the single-snapshot covariance matrix, the Toeplitz matrix is used to reconstruct the covariance matrix of the single-snapshot channel data and perform decorrelation processing;

[0067] Beam-domain MUSIC angle estimation module: This module takes the covariance matrix of the decorrelated single-snapshot channel data as input and uses the beam-domain MUSIC algorithm to perform super-resolution angle estimation of the coherent source angle.

[0068] The present invention also discloses a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the above-mentioned coherent signal source single-snapshot super-resolution angle estimation method is implemented.

[0069] The present invention also discloses a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned coherent signal source single-snapshot super-resolution angle estimation method is implemented.

[0070] Compared with the prior art, the advantages of the present invention are:

[0071] 1. This invention effectively improves the ability to distinguish coherent sources through decorrelation. Even when the angles between targets are very close, accurate angle estimation can be performed, resolving the inability of traditional methods to distinguish coherent sources and significantly improving the accuracy of DOA (direction of arrival) estimation.

[0072] 2. Using single-snapshot data for angle estimation not only improves processing speed but also reduces the system's computational burden, making it particularly suitable for applications with high real-time requirements. Single-snapshot processing enables the system to complete data processing in a shorter time, resulting in lower latency.

[0073] 3. Reconstructing the covariance matrix using the Toeplitz matrix effectively avoids the rank loss problem caused by signal correlation, thereby improving system stability and robustness. In particular, in situations with poor signal quality or severe multipath interference, the system can operate stably and reduce angle estimation errors.

[0074] 4. This invention can achieve super-resolution angle estimation with a relatively low number of array elements, adapting to radar array systems of varying sizes and offering strong application flexibility. Even with limited resources, it can still ensure high-precision angle estimation, demonstrating promising practical application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 Schematic diagram of the principle of an embodiment of the present invention;

[0076] Figure 2 This is a schematic diagram of single snapshot channel data selection according to an embodiment of the present invention;

[0077] Figure 3 : is the directional diagram of the coherent signal source TB-MUSIC method according to an embodiment of the present invention;

[0078] Figure 4 This is a diagram showing the angle estimation accuracy of the coherent signal source TB-MUSIC method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0079] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and examples.

[0080] like Figure 1 As shown, the present invention provides a Toeplitz matrix-based coherent source single snapshot super-resolution angle estimation (TB-MUSIC) method, comprising:

[0081] Step 1: Perform range processing and Doppler processing on the echo signal of the entire coherent integration period of the radar to obtain the range-Doppler-channel spectrum. Fix the range unit and Doppler unit where the target is located, extract the single-snapshot channel data from the CRD spectrum, and calculate the single-snapshot covariance matrix.

[0082] In a preferred embodiment, step one specifically includes:

[0083] Assuming that the radar's transmitted signal is a linear frequency modulation signal, the transmitted signal during the entire coherent accumulation period is:

[0084]

[0085] Where H is the number of pulses, T r is the pulse repetition period, f c is the carrier frequency, k = B / T is the frequency modulation slope, B is the bandwidth, and T is the pulse width. The number of points in each pulse is M = f s ·T r , f s is the sampling frequency, T / T r The duty cycle is set to 1 / 15.

[0086] The radar receiving signal is:

[0087]

[0088] Where τ is the target delay, which is related to the distance R and speed v:

[0089]

[0090] The antenna array uses a uniform linear array of N elements. Taking the leftmost antenna as the reference, the N×1 dimensional spatial steering matrix is:

[0091] a(θ)=[0,...,e j2π(N-1)dsinθ / λ ] T

[0092] Where λ is the wavelength, d = λ / 2 is the array element spacing, and θ is the target location.

[0093] Assume there are two coherent targets in space, with the same distance and speed, but different directions, and one with a strong amplitude and the other with a weak amplitude. The received signal is:

[0094] S r (t)=a(θ0)·S0·S r0 (t)+a(θ1)·S1·S r0 (t)+n(t)

[0095] Where S0 and S1 are the amplitudes of the two targets, S0>S1, θ0 is the angle of target 1, θ1 is the angle of target 2, and n(t) is Gaussian white noise.

[0096] The radar signal over the entire coherent integration period is transformed into three dimensions, with dimensions H·M·N, where H is the number of pulses, M is the number of points in a single pulse, and N is the number of array elements. The RD spectrum is plotted using the first channel of the simulated echo signal's CRD spectrum. This includes two coherent targets, one strong and one weak, positioned at 0° and 10°, respectively, with an angle difference of 10°. The array consists of eight elements.

[0097] The basic idea of distance processing is to perform H pulse compression calculations on H fast time-dimensional pulses through pulse compression, and determine the distance information through the peak value. Pulse compression is to set a matched filter, which is expressed as the deconvolution and conjugation of the transmitted signal:

[0098] y R (t) = S r (t)*h(t)

[0099]

[0100] Where * represents convolution. As can be seen from the above formula, the phase-frequency characteristics of the matched filter are opposite to those of the received signal. Through convolution, the signal's phase can be reduced to zero, achieving coherent accumulation. Conversely, the phase of noise is random, so only non-coherent accumulation can be achieved, and the peak value can be used to clearly determine the target's distance information.

[0101] The pulse compression process is implemented using the frequency domain method:

[0102] S r (w) = fft[S r (t)],H(w)=fft[h(t)]

[0103] Y R (w)=S r (w)·H(w),y R (t) = ifft[YR (w)]

[0104] The received signal and the matched filter can be transformed into the frequency domain through Fourier transform, multiplied and then transformed back into the time domain using inverse Fourier transform to complete the distance processing.

[0105] The basic idea of Doppler processing is to perform Fourier transform on the slow time dimension to obtain the range-Doppler-channel spectrum and clarify the distance information and velocity information.

[0106] y RD (t) = fft[y R (t)·w hamming ] / H

[0107] Among them, w hamming is a Hamming window of length H.

[0108] From the RD spectrum, two coherent targets are in the same range unit and Doppler unit. The range unit and Doppler unit where the target is located are fixed, and the N×1-dimensional single-snap channel data is extracted. The schematic diagram of selecting the single-snap channel data is as follows: Figure 2 As shown, calculate the covariance matrix of single snapshot channel data:

[0109] R=y RDC ·y RDC H

[0110] Among them, y RDC is a single snapshot channel data, R is the N×N dimensional covariance matrix, [·] H represents the conjugate transpose.

[0111] Step 2: Based on the obtained single-snapshot covariance matrix, the Toeplitz matrix is used to reconstruct the covariance matrix of the single-snapshot channel data to achieve decorrelation processing;

[0112] In a preferred embodiment, step 2 specifically includes:

[0113] Using the above covariance matrix for DOA estimation will not be able to distinguish between two coherent targets because the decorrelation operation is not performed. This is mainly due to the rank loss. If the two targets are uncorrelated, the covariance matrix is full rank, and its rows and columns are uncorrelated; but if the two targets are coherent, then linearly correlated rows or columns will appear in the covariance matrix, resulting in rank loss. Therefore, a single-snapshot channel data is selected, and the covariance matrix of the single-snapshot channel data is used to construct the Toeplitz matrix to achieve decorrelation processing and make the newly constructed covariance matrix full rank. The structure of the Toeplitz matrix is:

[0114]

[0115] Among them, a0~a n-1 is the first column of the covariance matrix of the single snapshot channel data, a0~a -(n-1) The first row of the covariance matrix of the single snapshot channel data is shown in Figure 2. Since the rank of the Toeplitz matrix is only related to the direction of arrival, the adverse effects of signal correlation on angle estimation are avoided, achieving decorrelation.

[0116] Step 3: Use the covariance matrix obtained after decorrelation processing to perform super-resolution angle estimation using the beam domain MUSIC algorithm.

[0117] In a preferred embodiment, step three specifically includes:

[0118] The MUSIC method in the beam domain converts the channel information into beam information, and then uses the MUSIC method to estimate the angle. The data is transformed into the beam domain using the conversion matrix, and the conversion matrix dimension is M×N B , which has the form:

[0119]

[0120] Among them, N B is the number of beams, is the angle of the composite beam.

[0121] The data is transformed into the beam domain using the covariance matrix and transformation matrix reconstructed by the Toeplitz matrix:

[0122]

[0123] Among them, R B is the covariance matrix in the beam domain, whose dimension is N B ×N B .

[0124] Using the covariance matrix in the beam domain, the angle is estimated using the MUSIC method. The eigenvalues and eigenvectors are obtained through eigenvalue decomposition. The eigenvectors are then rearranged in descending order of eigenvalues:

[0125] R B =UΣU H

[0126] sort(U)sort(Σ)

[0127] Among them, Σ is the eigenvalue, U is the eigenvector, and it is divided into signal subspace and noise subspace according to the number of sources. The first S columns with larger eigenvalues represent the signal subspace, and the last N columns with smaller eigenvalues represent the noise subspace. B - The S column represents the noise subspace.

[0128] Generate a spatial compensation vector:

[0129]

[0130] The noise subspace, transformation matrix and spatial compensation vector are used to find the spatial spectrum estimation coefficients:

[0131]

[0132] In summary, the present invention comprehensively utilizes the TB-MUSIC method to complete DOA estimation. By selecting single snapshot channel data from the CRD spectrum, the Toeplitz matrix decorrelation correction and the beam domain MUSIC method are combined to invent the TB-MUSIC method to complete the decorrelation operation of the coherent signal source and realize the angle estimation of two coherent signal sources, one strong and one weak. A uniform linear array of eight elements is selected in the simulation data, and the element spacing is half a wavelength. The two coherent targets in the simulation data differ by 10°. Digital beam synthesis cannot break through the angular resolution to obtain two peaks. When using the conventional MUSIC method, although the limitation of the angular resolution is broken through, the decorrelation cannot be completed, and the two targets are regarded as one target. By using the TB-MUSIC method in the present invention, not only can the limitation of the angular resolution be broken through, but the Toeplitz matrix can also be used to complete the decorrelation correction, and the beam MUSIC method is used for angle estimation to complete the angle estimation of two coherent targets, one strong and one weak. The angle estimation results are as follows. Figure 3 As shown. Figure 4 It can be seen that the angle estimation error of the TB-MUSIC method in the present invention is less than 1° when the channel signal-to-noise ratio is -20 dB.

[0133] In another embodiment of the present invention, a coherent source single-snapshot super-resolution angle estimation system based on a Toeplitz matrix is provided. The system can be used to implement the above-mentioned coherent source single-snapshot super-resolution angle estimation method, specifically including:

[0134] Pulse compression and Doppler processing module: used to perform pulse compression and Doppler processing on the radar echo signal of a complete coherent integration cycle, obtain the target's range and velocity information, obtain the CRD spectrum, fix the target's range unit and Doppler unit, extract the single-snapshot channel data from the CRD spectrum, and calculate the single-snapshot covariance matrix;

[0135] Toeplitz matrix reconstruction module: Based on the single-snapshot covariance matrix, the Toeplitz matrix is used to reconstruct the covariance matrix of the single-snapshot channel data and perform decorrelation processing;

[0136] Beam-domain MUSIC angle estimation module: This module takes the covariance matrix of the decorrelated single-snapshot channel data as input and uses the beam-domain MUSIC algorithm to perform super-resolution angle estimation of the coherent source angle.

[0137] In another embodiment of the present invention, a terminal device is provided, comprising a processor and a memory, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the coherent source single-snapshot super-resolution angle estimation method.

[0138] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory.

[0139] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the method for single-snapshot super-resolution angle estimation of a coherent signal source in the above embodiment; one or more instructions in the computer-readable storage medium are loaded and executed by the processor.

[0140] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0141] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0142] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0144] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the implementation methods of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific descriptions and embodiments. Those skilled in the art can make various other specific variations and combinations based on the technical teachings disclosed in the present invention without departing from the essence of the present invention, and such variations and combinations are still within the scope of protection of the present invention.

Claims

1. A method for super-resolution angle estimation of a coherent source single snapshot based on a Toeplitz matrix, comprising the following steps: S1: Perform pulse compression and Doppler processing on the radar echo signal of a complete coherent integration cycle to obtain the target's range and velocity information, obtain the range-Doppler-channel CRD spectrum, fix the target's range unit and Doppler unit, extract the single-snapshot channel data from the CRD spectrum, and calculate the single-snapshot covariance matrix; S2: Based on the obtained single-snapshot covariance matrix, the Toeplitz matrix is used to reconstruct the covariance matrix of the single-snapshot channel data to achieve decorrelation processing; S3: The covariance matrix of the decorrelated single snapshot channel data is used as input, and the angle of the coherent source is estimated by the beam-domain MUSIC algorithm.

2. The method for super-resolution angle estimation of a coherent source single snapshot according to claim 1, characterized in that: The pulse compression process processes the pulse signal through a matched filter to determine the target's distance information through the peak value; and the Doppler processing obtains the range-Doppler-channel spectrum through Fourier transform to clarify the target's distance and speed information.

3. The method for super-resolution angle estimation of a coherent source single snapshot according to claim 1, wherein: In S1, it is assumed that the radar's transmitted signal is a linear frequency modulation signal. The transmitted signal during the entire coherent accumulation period is: Where H is the number of pulses, T r is the pulse repetition period, f c is the carrier frequency, k = B / T is the frequency modulation slope, B is the bandwidth, T is the pulse width; the number of points in each pulse is M = f s ·T r , f s is the sampling frequency, T / T r The duty cycle is set to 1 / 15; The radar receiving signal is: Where τ is the target delay, which is related to the distance R and speed v: The antenna array uses a uniform linear array of N elements. Taking the leftmost antenna as the reference, the N×1 dimensional spatial steering matrix is: a(θ)=[0,...,e j2π(N-1)dsinθ / λ ] T Where λ is the wavelength, d = λ / 2 is the array element spacing, and θ is the target location; Assume there are two coherent targets in space, with the same distance and speed, but different directions, and one with a strong amplitude and the other with a weak amplitude. The received signal is: S r (t)=a(θ0)·S0·S r0 (t)+a(θ1)·S1·S r0 (t)+n(t) Where S0 and S1 are the amplitudes of the two targets, S0>S1, θ0 is the angle of target 1, θ1 is the angle of target 2, and n(t) is Gaussian white noise.

4. The method for super-resolution angle estimation of a coherent source single snapshot according to claim 3, wherein: The range Doppler processing in S1 includes: The radar signal of the entire coherent accumulation period is transformed into three dimensions with the dimension of H·M·N, where H is the number of pulses, M is the number of points of a single pulse, and N is the number of array elements. The basic idea of distance processing is to perform H pulse compression calculations on H fast time-dimensional pulses through pulse compression, and determine the distance information through the peak value. Pulse compression is to set a matched filter, which is the deconvolution and conjugation of the transmitted signal, specifically: y R (t)=S r (t)*h(t) Where * represents convolution calculation. As can be seen from the above formula, the phase-frequency characteristic of the matched filter is opposite to that of the received signal. Through convolution, the phase of the signal can be reduced to 0, achieving coherent accumulation. Conversely, the phase of the noise is random, so it can only complete non-coherent accumulation, and the peak value can be used to clearly determine the distance information of the target. The pulse compression process is implemented using the frequency domain method: S r (w)=fft[S r (t)],H(w)=fft[h(t)] Y R (w)=S r (w)·H(w),y R (t)=ifft[Y R (w)] The received signal and the matched filter can be transformed into the frequency domain through Fourier transform, multiplied and then transformed back into the time domain through inverse Fourier transform to complete the distance processing; Doppler processing is performed on the radar echo signal in S1: Obtain range-Doppler-channel spectrum to clarify distance and velocity information; y RD (t)=fft[y R (t)·w hamming ] / H Among them, w hamming is a Hamming window of length H; Extract N×1-dimensional single-snapshot channel data from the RD spectrum and calculate its covariance matrix: R=y RDC ·y RDC H Among them, y RDC is a single snapshot channel data, R is the N×N dimensional covariance matrix, [·] H represents the conjugate transpose.

5. The method for super-resolution angle estimation of a coherent source single snapshot according to claim 4, characterized in that: In S2, the Toeplitz matrix is used to perform decorrelation processing on the covariance matrix of the single snapshot channel data. The specific method is: The Toeplitz matrix is constructed using the covariance matrix of the single snapshot channel data, so that the newly constructed covariance matrix is full rank; The structure of the Toeplitz matrix is: Among them, a0~a n-1 is the first column of the covariance matrix of the single snapshot channel data, a0~a -(n-1) It is the first row of the covariance matrix of the single snapshot channel data.

6. The method for super-resolution angle estimation of a coherent source single snapshot according to claim 5, characterized in that: In S3, the covariance matrix obtained after the decorrelation processing is used to perform the beam domain MUSIC algorithm to perform super-resolution angle estimation. The specific method is: Convert channel information into beam information and transform data into beam domain using conversion matrix. The dimension of the conversion matrix is M×N. B , which has the form: Among them, N B is the number of beams, is the angle of the composite beam; Using the covariance matrix and transformation matrix reconstructed by the Toeplitz matrix, the data is transformed into the beam domain to obtain the covariance matrix in the beam domain: Among them, R B is the covariance matrix in the beam domain, whose dimension is N B ×N B ; Using the covariance matrix in the beam domain, the MUSIC method is used to estimate the angle and perform eigenvalue decomposition to obtain eigenvalues and eigenvectors. The eigenvectors are then rearranged in descending order of eigenvalues. The formula is as follows: R B =UΣU H sort(U)sort(Σ) Among them, Σ is the eigenvalue, U is the eigenvector, and it is divided into signal subspace and noise subspace according to the number of sources. The first S columns with larger eigenvalues represent the signal subspace, and the last N columns with smaller eigenvalues represent the noise subspace. B - The S column represents the noise subspace; Generate a spatial compensation vector: The noise subspace, transformation matrix and spatial compensation vector are used to find the spatial spectrum estimation coefficients: Where θ is the azimuth angle of the target, n is the number of array elements, d is the element spacing, and λ is the wavelength.

7. A coherent source single-snapshot super-resolution angle estimation system based on the Toeplitz matrix, characterized by: The system can be used to implement the coherent source single-snapshot super-resolution angle estimation method according to any one of claims 1 to 6, specifically comprising: Pulse compression and Doppler processing module: used to perform pulse compression and Doppler processing on the radar echo signal of a complete coherent integration cycle, obtain the target's range and velocity information, obtain the CRD spectrum, fix the target's range unit and Doppler unit, extract the single-snapshot channel data from the CRD spectrum, and calculate the single-snapshot covariance matrix; Toeplitz matrix reconstruction module: Based on the single-snapshot covariance matrix, the Toeplitz matrix is used to reconstruct the covariance matrix of the single-snapshot channel data and perform decorrelation processing; Beam-domain MUSIC angle estimation module: This module takes the covariance matrix of the decorrelated single-snapshot channel data as input and uses the beam-domain MUSIC algorithm to perform super-resolution angle estimation of the coherent source angle.

8. A computer device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for super-resolution angle estimation of a single snapshot of a coherent signal source according to one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method for super-resolution angle estimation of a coherent signal source single snapshot according to one of claims 1 to 6 is implemented.

Images