DOA estimation method of 1-bit low-complexity array structure based on atom norm minimization

By using a 1-bit low-complexity array structure with atomic norm minimized atomic norm and a simulated phase shifter network, the problems of phase blur caused by HAD architecture and the reduction of accuracy of 1-bit ADC are solved, and efficient DOA estimation is achieved. Reduces cost and complexity and improves estimation accuracy.

CN118938118AActive Publication Date: 2024-11-12NANJING UNIV OF POSTS & TELECOMM

Patent Information

Application Number
CN202410976507.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-20
Publication Date
2024-11-12
Estimated Expiration
2044-07-20

AI Technical Summary

Technical Problem

Existing HAD architectures cause phase blur problems, and using a 1-bit ADC in large-scale/hyperscale arrays reduces DOA estimation accuracy.

Method used

Using a 1-bit low-complexity array structure based on atomic norm minimization, a hybrid analog digital structure and a 1-bit ADC connected by two-stage partially, an analog phase shifter network is designed to solve the problem of signal dimensionality reduction and atomic norm minimization. The unquantized noiseless component is restored to obtain an angle estimate.

Benefits of technology

It effectively avoids phase blur and beam scanning problems, reduces hardware cost and complexity, and improves DOA estimation accuracy, avoids grid mismatch effect and covariance matrix reconstruction error in 1-bit ADC scenario.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938118A_ABST
    Figure CN118938118A_ABST
Patent Text Reader

Abstract

The invention provides a DOA (Direction of Arrival) estimation method of a 1-bit low-complexity array structure based on atom norm minimization, which comprises the following steps of: constructing a hybrid analog digital structure with two-stage part connection at a receiving end, and using a 1-bit ADC (Analog to Digital Converter) at an ADC (Analog to Digital Converter) part; the method comprises the following steps of: receiving a signal by utilizing a TL-SC HAD array of a framework and modeling; designing a phase shifter network initial value and obtaining an output signal; dimensionality reduction is carried out according to the output signal, an atomic norm minimization problem is established, and an angle estimation value is obtained according to the recovered noise-free component; and updating the network weight of the phase shifter by using the obtained angle estimation value, and repeatedly performing angle estimation until the maximum number of iterations or satisfactory estimation precision is reached, thereby obtaining a final angle estimation value. According to the method, the hardware cost and power consumption of the array can be reduced, and meanwhile, higher estimation precision can be realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of signal processing, and specifically is a DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization, which can be used in large-scale / ultra-large-scale array direction of arrival estimation scenarios. Background Art

[0002] Direction-of-Arrival (DOA) estimation is an important branch of array signal processing. It uses array antennas to receive spatial signals and processes the received signals through statistical signal processing techniques and various optimization methods to restore the direction information of the incident signal. It has a wide range of applications in radar, sonar, voice and wireless communications.

[0003] There are generally two ways to improve the accuracy of DOA estimation. One is to develop a high-resolution DOA estimation algorithm, and the other is to optimize the array and increase the array size to improve the estimation accuracy. In recent years, large-scale / ultra-large-scale antenna arrays that deploy a large number of elements can bring ultra-high estimation accuracy to DOA estimation and have received increasing attention. However, due to the large number of antennas, the same number of RF links is required to match them, including expensive devices such as analog to digital converters (ADCs), which greatly increases hardware costs and energy consumption.

[0004] In order to solve the problem of high hardware cost and energy consumption, the hybrid analog and digital (HAD) array structure using limited RF chains is widely adopted. In the sub-connected (SC) HAD structure, each antenna is connected to a phase shifter, and multiple phase shifters are connected to an RF chain, reducing the number of required RF chains. In order to further reduce the complexity and power consumption of the array structure, the use of low-precision ADC in the ADC part has attracted widespread attention. Among them, 1-bit quantization is an extreme case. The quantization method is to compare the received signal with a preset threshold in advance, output 1 when it is greater than the threshold, and output negative 1 when it is less than or equal to the threshold. The threshold is usually set to 0, and can also be set to a time-varying threshold. After quantization, the 1-bit ADC only retains the symbol information of the received data, reducing the amount of information transmission and storage, and can improve the feasibility in large-scale / ultra-large-scale array scenarios.

[0005] However, the SC HAD structure will bring about phase ambiguity, which requires multiple time slots to eliminate this phase ambiguity. As the number of subarray antennas increases, the DOA measurement delay will increase linearly. At the same time, after 1-bit quantization, only the symbol information of the received signal is retained, which brings challenges to DOA estimation under the SC HAD structure.

[0006] The above problems are issues that should be considered and solved in the DOA estimation process of 1-bit low-complexity array structure based on atomic norm minimization. Summary of the invention

[0007] The purpose of the present invention is to provide a DOA estimation method of a 1-bit low-complexity array structure based on atomic norm minimization to solve the problems in the prior art that the existing HAD architecture will cause phase ambiguity and the use of a 1-bit ADC will reduce the estimation accuracy.

[0008] The technical solution of the present invention is:

[0009] A DOA estimation method for a 1-bit quantized low-complexity array structure based on atomic norm minimization comprises the following steps:

[0010] S1. A 1-bit hybrid analog-digital array is constructed at the receiving end, wherein the array structure adopts a two-stage partially connected hybrid analog-digital structure, and the ADC part uses a 1-bit ADC;

[0011] S2, using the constructed array to receive the incident signal and model it;

[0012] S3, design the initial value of the analog phase shifter network and obtain the output signal;

[0013] S4, reducing the dimension of the output signal Y of the L snapshot;

[0014] S5. Establish an atomic norm minimization problem and solve it to recover the unquantized noise-free component;

[0015] S6. Obtain an angle estimate based on the restored noise-free component z

[0016] S7, using the result obtained in step S6 The original phase in step S3 is replaced, and steps S4 to S6 are repeated until the maximum number of iterations or a satisfactory estimation accuracy is reached, thereby obtaining a final angle estimation value.

[0017] Furthermore, in step S1, a 1-bit hybrid analog-digital array is constructed at the receiving end, specifically,

[0018] S11, the hybrid analog digital array includes N array elements, which are divided into K sub-arrays;

[0019] S12. The first subarray of the K subarrays contains M1 array elements, each of which is connected to an RF chain. Each of the remaining K-1 subarrays contains M array elements and adopts a SC HAD structure, where M=M1+1.

[0020] S13. The output of each RF chain can be regarded as the output of a virtual array element. The total number of RF chains is N. RF =M1+K-1, and a 1-bit ADC is deployed in the ADC part.

[0021] Furthermore, in step S2, the incident signal is received and modeled using the constructed array, specifically,

[0022] S21. Assuming that the far-field narrowband signal is incident at an azimuth angle θ, the array receiving signal at time t can be expressed as

[0023]

[0024] in The antenna receives the signal. is a signal source of unknown direction, j is an imaginary number, f c is the carrier frequency, w(t) is the additive white Gaussian noise, a(θ) is the array manifold vector, expressed as a(θ) = [1,e jπsin(θ) ,…,e j π(N-1)sin(θ) ] T , where e is a natural constant, N is the number of array elements, (·) T represents transpose;

[0025] S22, after phase calibration by analog phase shifter network, the received signal is expressed as

[0026]

[0027] in,(·) H represents the conjugate transpose, V A is the analog beamforming matrix, expressed as in represents the M1×M1 dimensional identity matrix, v A,k (k≥2) represents the simulated beamforming vector of the kth subarray, expressed as where α k,m ,m=1,…,M is the phase of the mth phase shifter in the kth subarray;

[0028] S23, RF signal After the parallel RF chain, it is down-converted to baseband signal This signal is generated by a 1-bit ADC

[0029]

[0030] in, represents the complex-valued element quantization function composed of two sign functions sign(x), expressed as in, It means to take the real part of the complex number x. It means to take the imaginary part of the complex number x, and the sign function is The array output signal is expressed as

[0031]

[0032] Wherein, y(l) is the output signal, l=1, 2, ..., L, and L is the number of snapshots.

[0033] Further, in step S3, the initial value of the analog phase shifter network is designed and the output signal is obtained, specifically,

[0034] S31, using the quantized data of the first subarray, an initial estimated value θ0 can be obtained according to the existing one-bit MUSIC method. In an ideal case, θ0=θ;

[0035] S32, when k ≥ 2, use this angle value to design the phase of the phase shifter network, that is,

[0036]

[0037] S33, the output signal of the kth (k≥2) subarray is

[0038]

[0039] where w k (t) is the additive white Gaussian noise of the kth (k ≥ 2) subarray. The set of virtual array element positions corresponding to the K RF chain outputs is in, For nth RF The index of the virtual array elements, the output signals of the K sub-arrays are written in vector form as

[0040]

[0041] in, is the steering vector corresponding to the set Ω, and D is the gain matrix of the subarray, specifically: in, To select the matrix, In the jth row of Γ, only Ω j The value of the corresponding position is 1, and the other positions are 0;

[0042] S34, quantize the signal by 1 bit to obtain an output signal The output signal of the L snapshot is Y=[y(1),...,y(L)].

[0043] Further, in step S4, the output signal Y of the L snapshot is reduced in dimension, specifically, Y is subjected to singular value decomposition to obtain Y=U Y L Y V Y , let d K =[1,0,...,0] to preserve the signal subspace, and get

[0044] Furthermore, in step S5, an atomic norm minimization problem is established to solve and recover the unquantized noise-free component, specifically,

[0045] S51. For the unquantized noise-free component that we want to recover, z=a(θ)s, we establish an atomic set: The atomic norm of z is expressed as

[0046] S52. Establish the atomic norm minimization problem:

[0047]

[0048] Where γ is the regularization parameter. Solve to get the noise-free component z.

[0049] Further, in step 52, the atomic norm minimization problem is solved to obtain the noise-free component z, specifically,

[0050] S521. The atomic norm minimization problem can be equivalently expressed as the following optimization problem:

[0051]

[0052] Where x is a free variable, u1 is the first element of vector u, T(u) is the Toeplitz matrix with u as the first column, and z R and z I denote the real and imaginary parts of z respectively;

[0053] S522, the constructed optimization problem is solved using the ADMM method, and the optimization problem is rewritten as

[0054]

[0055] in, and Respectively The real and imaginary parts of , U and α are auxiliary variables;

[0056] S523. The augmented Lagrangian function of the optimization problem is:

[0057]

[0058] in, is the Lagrange multiplier, let in ρ is the penalty factor;

[0059] S524, the update step of the i+1th time in ADMM is obtained from the augmented Lagrangian function:

[0060]

[0061]

[0062] in, B=diag([N s ,2(N s -1),…,2]), e1=[1,0,…,0] T , and considering U ≥ 0 and α ≥ 0, perform eigenvalue decomposition on the corresponding update formula and set all negative eigenvalues ​​to zero. ADMM iterates until convergence to obtain the noise-free component z = z R +jz I .

[0063] Further, in step S6, the angle estimation value is obtained according to the restored noise-free component z. Specifically, the angle estimation method uses the root-MUSIC algorithm to calculate the covariance matrix R of z z =zz H , to get R z Perform eigenvalue decomposition and sort the eigenvalues ​​in order of size to obtain the signal subspace and noise subspace, and construct the polynomial in, Find the root of the polynomial and take the root closest to the unit circle pass Get the angle estimate.

[0064] The beneficial effects of the present invention are:

[0065] 1. This DOA estimation method based on a 1-bit low-complexity array structure with atomic norm minimization, through a TL-SC HAD array architecture, fully utilizes the characteristics of the nested array, can effectively avoid the problems of phase ambiguity and beam scanning, and uses a 1-bit ADC in the ADC part, which can further reduce the cost and complexity, and to a certain extent guarantee the DOA estimation accuracy.

[0066] Second, this DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization is not affected by the grid mismatch effect, and can avoid the error between the normalized covariance matrix and the unquantized covariance matrix obtained by the covariance matrix reconstruction method in the 1-bit ADC scenario, which can effectively improve the estimation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a flowchart of a DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to an embodiment of the present invention;

[0068] Figure 2 is a schematic diagram of the structure of a 1-bit TL-SC HAD array architecture in an embodiment;

[0069] Figure 3 is a schematic diagram of performance comparison between a DOA estimation method of a 1-bit low-complexity array structure based on atomic norm minimization described in an embodiment and other existing methods; DETAILED DESCRIPTION

[0070] The preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0071] Example

[0072] A DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization, specifically, comprising the following steps:

[0073] S1. A 1-bit hybrid analog-digital array is constructed at the receiving end, wherein the array structure adopts a two-stage partially connected hybrid analog-digital structure, and the ADC part uses a 1-bit ADC;

[0074] S11, the hybrid analog digital array includes N array elements, which are divided into K sub-arrays;

[0075] S12. The first subarray of the K subarrays contains M1 array elements, each of which is connected to an RF chain. Each of the remaining K-1 subarrays contains M array elements and adopts a SC HAD structure, where M=M1+1.

[0076] S13. The output of each RF chain can be regarded as the output of a virtual array element. The total number of RF chains is N. RF =M1+K-1, and a 1-bit ADC is deployed in the ADC part.

[0077] S2, using the constructed array to receive the incident signal and model it;

[0078] S21. Assuming that the far-field narrowband signal is incident at an azimuth angle θ, the array receiving signal at time t can be expressed as

[0079]

[0080] in The antenna receives the signal. is a signal source of unknown direction, j is an imaginary number, f c is the carrier frequency, w(t) is the additive white Gaussian noise, a(θ) is the array manifold vector, expressed as a(θ) = [1,e jπsin(θ) ,...,e j π(N-1)sin(θ) ] T , where e is a natural constant, N is the number of array elements, (·) T represents transpose;

[0081] S22, after phase calibration by analog phase shifter network, the received signal is expressed as

[0082]

[0083] in,(·) H represents the conjugate transpose, V A is the analog beamforming matrix, expressed as in represents the M1×M1 dimensional identity matrix, v A,k (k≥2) represents the simulated beamforming vector of the kth subarray, expressed as where α k,m ,m=1,…,M is the phase of the mth phase shifter in the kth subarray;

[0084] S23, RF signal After the parallel RF chain, it is down-converted to baseband signal This signal is generated by a 1-bit ADC

[0085]

[0086] in, represents the complex-valued element quantization function composed of two sign functions sign(x), expressed as in, It means to take the real part of the complex number x. It means to take the imaginary part of the complex number x, and the sign function is The array output signal is expressed as

[0087]

[0088] Wherein, y(l) is the output signal, l=1, 2, ..., L, and L is the number of snapshots.

[0089] S3, design the initial value of the analog phase shifter network and obtain the output signal;

[0090] S31, using the quantized data of the first subarray, an initial estimated value θ0 can be obtained according to the existing one-bit MUSIC method. In an ideal case, θ0=θ;

[0091] S32, when k ≥ 2, use this angle value to design the phase of the phase shifter network, that is,

[0092]

[0093] S33, the output signal of the kth (k≥2) subarray is

[0094]

[0095] where w k (t) is the additive white Gaussian noise of the kth (k ≥ 2) subarray. The set of virtual array element positions corresponding to the K RF chain outputs is in, For nth RF The index of the virtual array elements, the output signals of the K sub-arrays are written in vector form as

[0096]

[0097] in, is the steering vector corresponding to the set Ω, and D is the gain matrix of the subarray, specifically: in, To select the matrix, In the jth row of Γ, only Ω j The value of the corresponding position is 1, and the other positions are 0;

[0098] S34, quantize the signal by 1 bit to obtain an output signal The output signal of the L snapshot is Y=[y(1),...,y(L)].

[0099] S4, reducing the dimension of the output signal Y of the L snapshot;

[0100] In step S4, the output signal Y of the L snapshot is reduced in dimension. Specifically, Y is subjected to singular value decomposition to obtain Y=U Y L Y V Y , let d K =[1,0,...,0] to preserve the signal subspace, and get

[0101] S5. Establish an atomic norm minimization problem and solve it to recover the unquantized noise-free component;

[0102] S51. For the unquantized noise-free component that we want to recover, z=a(θ)s, we establish an atomic set: The atomic norm of z is expressed as

[0103] S52. Establish the atomic norm minimization problem:

[0104]

[0105] Where γ is the regularization parameter. Solve to get the noise-free component z.

[0106] In step 52, the atomic norm minimization problem is solved to obtain the noise-free component z, specifically,

[0107] S521. The atomic norm minimization problem can be equivalently expressed as the following optimization problem:

[0108]

[0109] Where x is a free variable, u1 is the first element of vector u, T(u) is the Toeplitz matrix with u as the first column, and z R and z I denote the real and imaginary parts of z respectively;

[0110] S522, the constructed optimization problem is solved using the ADMM method, and the optimization problem is rewritten as

[0111]

[0112] in, and Respectively The real and imaginary parts of , U and α are auxiliary variables;

[0113] S523. The augmented Lagrangian function of the optimization problem is:

[0114]

[0115] in, is the Lagrange multiplier, let in ρ is the penalty factor;

[0116] S524, the update step of the i+1th time in ADMM is obtained from the augmented Lagrangian function:

[0117]

[0118]

[0119] in, B=diag([N s ,2(N s -1),…,2), e1=[1,0,…,0] T , and considering U ≥ 0 and α ≥ 0, perform eigenvalue decomposition on the corresponding update formula and set all negative eigenvalues ​​to zero. ADMM iterates until convergence to obtain the noise-free component z = z R +jz I .

[0120] S6. Obtain an angle estimate based on the restored noise-free component z

[0121] In step S6, the angle estimation value is obtained according to the restored noise-free component z Specifically, the angle estimation method uses the root-MUSIC algorithm to calculate the covariance matrix R of z z =zz H , to get R z Perform eigenvalue decomposition and sort the eigenvalues ​​in order of size to obtain the signal subspace and noise subspace, and construct the polynomial in, Find the root of the polynomial and take the root closest to the unit circle pass Get the angle estimate.

[0122] S7, using the result obtained in step S6 The original phase in step S3 is replaced, and steps S4 to S6 are repeated until the maximum number of iterations or a satisfactory estimation accuracy is reached, thereby obtaining a final angle estimation value.

[0123] This DOA estimation method based on a 1-bit low-complexity array structure with atomic norm minimization, through a TL-SC HAD array architecture, fully utilizes the characteristics of the nested array, can effectively avoid the problems of phase ambiguity and beam scanning, and uses a 1-bit ADC in the ADC part, which can further reduce the cost and complexity and ensure the DOA estimation accuracy to a certain extent.

[0124] This DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization is not affected by the grid mismatch effect, and can avoid the error between the normalized covariance matrix and the unquantized covariance matrix obtained by the covariance matrix reconstruction method in the 1-bit ADC scenario, which can effectively improve the estimation accuracy.

[0125] The effects of the present invention are further described below in conjunction with a simulation example.

[0126] Simulation example: A far-field narrowband signal is incident at a direction of 10°. The array contains 120 elements, which are divided into 11 sub-arrays. The first sub-array has 10 elements and 200 snapshots.

[0127] The DOA estimation method of the embodiment based on the 1-bit low-complexity array structure minimization of the atomic norm is compared with the estimation performance of other existing methods. Figure 3 As shown. Among them, the covariance reconstruction method is a 1-bit estimation method that uses the inverse sine rule to reconstruct the covariance matrix; the infinite precision quantization estimation method means that the ADC part in the array structure uses an infinite precision ADC, and the estimation method is a subspace estimation method based on a synthetic array; in the nested array 1-bit estimation method, the structure of the nested array is the same as the virtual array structure regarded as the RF chain output, and the angle estimation adopts a 1-bit estimation method of covariance matrix reconstruction. Figure 3 It can be seen that the embodiment method can achieve an estimation performance comparable to that of the infinite precision quantization estimation method under low signal-to-noise ratio conditions. Although it is greatly affected by the quantization noise at high signal-to-noise ratios, it can still achieve a higher estimation accuracy. The estimation accuracy can be better than the covariance reconstruction method and the nested array 1-bit estimation method. The above results reflect the effectiveness of the method proposed in the present invention.

[0128] In summary, the embodiment method combines the nested array concept with the HAD array structure and uses a 1-bit ADC in the ADC part, which can effectively reduce hardware cost and complexity. At the same time, the embodiment method can also achieve higher estimation accuracy.

[0129] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization, characterized in that: The following steps are included: S1. A 1-bit hybrid analog-digital array is constructed at the receiving end, wherein the array structure adopts a two-stage partially connected hybrid analog-digital structure, and the ADC part uses a 1-bit ADC; S2, using the constructed array to receive the incident signal and model it; S3, design the initial value of the analog phase shifter network and obtain the output signal; S4, reducing the dimension of the output signal Y of the L snapshot; S5. Establish an atomic norm minimization problem and solve it to recover the unquantized noise-free component; S6. Obtain an angle estimate based on the restored noise-free component z S7, using the result obtained in step S6 The original phase in step S3 is replaced, and steps S4 to S6 are repeated until the maximum number of iterations or a satisfactory estimation accuracy is reached, thereby obtaining a final angle estimation value.

2. The DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to claim 1, characterized in that: In step S1, a 1-bit hybrid analog-digital array is constructed at the receiving end, specifically, S11, the hybrid analog digital array includes N array elements, which are divided into K sub-arrays; S12. The first subarray of the K subarrays contains M1 array elements, each of which is connected to an RF chain. Each of the remaining K-1 subarrays contains M array elements and adopts a SC HAD structure, where M=M1+1. S13. The output of each RF chain can be regarded as the output of a virtual array element. The total number of RF chains is N. RF =M1+K-1, and a 1-bit ADC is deployed in the ADC part.

3. The DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to claim 1, characterized in that: In step S2, the incident signal is received and modeled using the constructed array, specifically, S21. Assuming that the far-field narrowband signal is incident at an azimuth angle θ, the array receiving signal at time t can be expressed as in The antenna receives the signal. is a signal source of unknown direction, j is an imaginary number, f c is the carrier frequency, w(t) is the additive white Gaussian noise, a(θ) is the array manifold vector, expressed as a(θ) = [1,e jπsin(θ) ,…,e jπ(N-1)sin(θ) ] T , where e is a natural constant, N is the number of array elements, (·) T represents transpose; S22, after phase calibration by analog phase shifter network, the received signal is expressed as in,(·) H represents the conjugate transpose, V A is the analog beamforming matrix, expressed as in represents the M1×M1 dimensional identity matrix, v A,k (k≥2) represents the simulated beamforming vector of the kth subarray, expressed as where α k,m ,m=1,…,M is the phase of the mth phase shifter in the kth subarray; S23, RF signal After the parallel RF chain, it is down-converted to baseband signal This signal is generated by a 1-bit ADC in, represents the complex-valued element quantization function composed of two sign functions sign(x), expressed as in, It means to take the real part of the complex number x. It means to take the imaginary part of the complex number x, and the sign function is The array output signal is expressed as Wherein, y(l) is the output signal, l=1, 2, ..., L, and L is the number of snapshots.

4. The DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to claim 1, characterized in that: In step S3, the initial value of the analog phase shifter network is designed and the output signal is obtained, specifically, S31, using the quantized data of the first subarray, an initial estimated value θ0 can be obtained according to the existing one-bit MUSIC method. In an ideal case, θ0=θ; S32, when k ≥ 2, use this angle value to design the phase of the phase shifter network, that is, S33, the output signal of the kth (k≥2) subarray is where w k (t) is the additive white Gaussian noise of the kth (k ≥ 2) subarray. The set of virtual array element positions corresponding to the K RF chain outputs is in, For nth RF The index of the virtual array elements, the output signals of the K sub-arrays are written in vector form as in, is the steering vector corresponding to the set Ω, and D is the gain matrix of the subarray, specifically: in, To select the matrix, In the jth row of Γ, only Ω j The corresponding position has a value of 1, and the remaining positions have a value of 0. S34, quantize the signal by 1 bit to obtain an output signal The output signal of the L snapshot is Y=[y(1),…,y(L)].

5. The DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to claims 1-4, characterized in that: In step S4, the output signal Y of the L snapshot is reduced in dimension. Specifically, Y is subjected to singular value decomposition to obtain Y=U Y L Y V Y , let d K =[1,0,…,0] is used to retain the signal subspace, and we get 6. The DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to claim 1, characterized in that: In step S5, an atomic norm minimization problem is established to solve and recover the unquantized noise-free component. Specifically, S51. For the unquantized noise-free component that we want to recover, z=a(θ)s, we establish an atomic set: The atomic norm of z is expressed as S52. Establish the atomic norm minimization problem: Where γ is the regularization parameter. Solve to get the noise-free component z.

7. The DOA estimation method of a 1-bit low-complexity array structure based on atomic norm minimization according to claim 6, characterized in that: In step 52, the atomic norm minimization problem is solved to obtain the noise-free component z, specifically, S521. The atomic norm minimization problem can be equivalently expressed as the following optimization problem: Where x is a free variable, u1 is the first element of vector u, T(u) is the Toeplitz matrix with u as the first column, and z R and z I denote the real and imaginary parts of z respectively; S522, the constructed optimization problem is solved using the ADMM method, and the optimization problem is rewritten as U≥0 α≥0 in, and Respectively The real and imaginary parts of , U and α are auxiliary variables; S523. The augmented Lagrangian function of the optimization problem is: in, is the Lagrange multiplier, let in ρ is the penalty factor; S524, the update step of the i+1th time in ADMM is obtained from the augmented Lagrangian function: in, B = diag(]N s ,2(N s -1),…,2]), e1=[1,0,…,0] T , and considering U ≥ 0 and α ≥ 0, perform eigenvalue decomposition on the corresponding update formula and set all negative eigenvalues ​​to zero. ADMM iterates until convergence to obtain the noise-free component z = z R +jz I .

8. The DOA estimation method for a 1-bit low-complexity array structure based on atomic norm minimization according to claim 1, characterized in that: In step S6, the angle estimation value is obtained according to the restored noise-free component z Specifically, the angle estimation method uses the root-MUSIC algorithm to calculate the covariance matrix R of z z =zz H , to get R z Perform eigenvalue decomposition and sort the eigenvalues ​​in order of size to obtain the signal subspace and noise subspace, and construct the polynomial in, Find the root of the polynomial and take the root closest to the unit circle pass Get the angle estimate.

Citation Information

Patent Citations

  • Meshless single-bit DOA estimation method based on nested crossed dipole array

    CN112363110A

  • Atomic norm minimization-based dimensionality-reducible two-dimensional out-of-lattice DOA estimation method

    CN113673317A

  • Grid-free single-bit DOA estimation method based on nested crossed dipole array

    WO2022110833A1

Cited By

  • Radar target parameter estimation method and device, electronic equipment, medium and product

    CN119738795A

  • HDS architecture-based DOA estimation method in THz super-large scale MIMO system

    CN120263236A