WO2023137812A1 - 基于虚拟域张量填充的互质面阵二维波达方向估计方法 - Google Patents

基于虚拟域张量填充的互质面阵二维波达方向估计方法 Download PDF

Info

Publication number
WO2023137812A1
WO2023137812A1 PCT/CN2022/076430 CN2022076430W WO2023137812A1 WO 2023137812 A1 WO2023137812 A1 WO 2023137812A1 CN 2022076430 W CN2022076430 W CN 2022076430W WO 2023137812 A1 WO2023137812 A1 WO 2023137812A1
Authority
WO
WIPO (PCT)
Prior art keywords
tensor
array
virtual
axis
virtual domain
Prior art date
Application number
PCT/CN2022/076430
Other languages
English (en)
French (fr)
Inventor
郑航
周成伟
史治国
王勇
陈积明
Original Assignee
浙江大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 浙江大学 filed Critical 浙江大学
Priority to US17/920,401 priority Critical patent/US20240210510A1/en
Publication of WO2023137812A1 publication Critical patent/WO2023137812A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00Computer-aided design [CAD]
    • G06F30/20Design optimisation, verification or simulation
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/02Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using radio waves
    • G01S3/14Systems for determining direction or deviation from predetermined direction
    • G01S3/143Systems for determining direction or deviation from predetermined direction by vectorial combination of signals derived from differently oriented antennae
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/02Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using radio waves
    • G01S3/74Multi-channel systems specially adapted for direction-finding, i.e. having a single antenna system capable of giving simultaneous indications of the directions of different signals
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/02Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using radio waves
    • G01S3/04Details
    • G01S3/043Receivers
    • HELECTRICITY
    • H01ELECTRIC ELEMENTS
    • H01QANTENNAS, i.e. RADIO AERIALS
    • H01Q21/00Antenna arrays or systems
    • H01Q21/06Arrays of individually energised antenna units similarly polarised and spaced apart
    • H01Q21/061Two dimensional planar arrays
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00Details relating to CAD techniques
    • G06F2111/10Numerical modelling

Definitions

  • the invention belongs to the technical field of array signal processing, and in particular relates to a statistical signal processing technology based on second-order high-dimensional statistics in a sparse array virtual domain, in particular to a two-dimensional DOA estimation method for a coprime array based on virtual domain tensor filling, which can be used for target positioning.
  • the coprime array As a sparse array with a systematic structure, the coprime array has the advantages of large aperture and high resolution, and can break through the performance bottleneck of traditional uniform array DOA estimation in terms of estimation performance and cost. Since the sparsely arranged elements of the coprime array do not meet the Nyquist sampling rate, in order to realize the DOA estimation of the Nyquist matching, the common practice is to calculate the second-order statistics of the received signal of the coprime array to construct an augmented non-continuous virtual array, and extract the continuous part from it to realize the Nyquist matching process based on the second-order equivalent signal in the virtual domain. Furthermore, in order to make full use of all the discontinuous virtual array elements, the existing method fills the discontinuous virtual array to improve the performance of DOA estimation.
  • the above method usually represents the received signal as a vector, and derives the second-order equivalent signal in the virtual domain by vectorizing the covariance matrix of the received signal; in the scenario of deploying a coprime array, since the received signal of the coprime array covers two-dimensional spatial information, this vectorized signal processing method destroys the original spatial information structure of the received signal of the coprime array, which will cause serious performance loss.
  • tensor As a multi-dimensional data type, has been applied in the field of array signal processing to characterize received signals covering multi-dimensional spatial information, and perform feature analysis and effective information extraction to achieve high-precision and high-resolution DOA estimation.
  • the augmented multi-dimensional non-continuous virtual array will have a whole patch of holes, resulting in a piece of missing elements in the corresponding virtual domain tensor.
  • the traditional tensor filling method applied to image restoration is based on the random distribution of missing elements in the tensor, so it cannot effectively fill the virtual domain tensor. Therefore, for the virtual domain tensor model of the coprime area array, how to effectively use all the discontinuous virtual domain tensor statistical information to achieve high-precision and high-resolution two-dimensional DOA estimation is still an urgent problem to be solved.
  • the purpose of the present invention is to solve the problems of the loss of multi-dimensional signal structure information and the inability to fully utilize virtual domain statistics existing in existing methods, and propose a two-dimensional direction of arrival estimation method based on virtual domain tensor filling of coprime arrays, which provides a feasible idea and an effective solution for making full use of all discontinuous virtual domain tensor statistics corresponding to coprime arrays to realize high-precision and high-resolution two-dimensional DOA estimation of Nyquist matching.
  • a method for estimating a two-dimensional direction of arrival of a coprime array based on virtual domain tensor filling comprising the following steps:
  • s k [s k,1 ,s k,2 ,...,s k,T ] T is the multi-snapshot sampling signal waveform corresponding to the kth incident signal source
  • [ ] T represents the transpose operation
  • the vector outer product represents the vector outer product
  • the signal source of is expressed as:
  • a two-dimensional augmented virtual plane array along the x-axis and y-axis direction is constructed, represents the Kronecker product; therefore, corresponds to a size of The non-continuous virtual array of The non-continuous virtual array Contains holes in the entire row and column, that is, missing elements;
  • Subscripts that will have the same s y index subscript Superimpose in the fourth dimension to obtain L y four-dimensional tensors whose dimensions are P x ⁇ P y ⁇ 2 ⁇ L x ; further, superpose these L y four-dimensional tensors in the fifth dimension to obtain a five-dimensional virtual domain tensor
  • This five-dimensional virtual domain tensor Covers spatial angle information in the x-axis and y-axis directions, spatial mirror transformation information, and spatial translation information in the x-axis and y-axis directions; defines a set of dimensions then pass Dimensions are combined to obtain a three-dimensional reconstructed virtual domain tensor
  • the three dimensions of represent spatial angle information, spatial translation information, and spatial mirror transformation information respectively.
  • the original virtual domain tensor Slices of missing elements in are randomly distributed to on the three spatial dimensions covered;
  • the optimization variable is the padded virtual domain tensor corresponding to the virtual uniform cubic array
  • the tensor kernel norm express A collection of positional indices of non-missing elements in , Indicates tensors in Mapping on ; Since the kernel norm is a convex function, the virtual field tensor filling problem based on the minimization of the tensor kernel norm is a solvable convex optimization problem. Solving the convex optimization problem, we get
  • step (2) By computing the tensor and The cross-correlation statistics of are approximated, that is, the sampling cross-correlation tensor
  • step (7) the filled virtual domain tensor Perform canonical polyadic decomposition to get the factor vector and then parameter and from and Extracted as:
  • ⁇ ( ⁇ ) represents the operation of taking the argument of a complex number
  • a (a) represents the ath element of a vector a; here, according to and ⁇ 1 ⁇ [1,P x P y -1] and ⁇ 2 ⁇ [1,L x L y -1] respectively satisfy mod( ⁇ 1 ,P x ) ⁇ 0 and mod( ⁇ 2 ,P y ) ⁇ 0, and ⁇ 1 ⁇ [1,P x P y -P x ], ⁇ 2 ⁇ [1,L x L y -L x ], mod( ⁇ ) represents the remainder operation; Parameter ( ⁇ k , v k ) and two-dimensional direction of arrival The relationship between the two-dimensional direction of arrival estimation is obtained
  • the closed-form solution of is:
  • the present invention has the following advantages:
  • the present invention derives the augmented discontinuous virtual array based on the cross-correlation tensor, and utilizes the mirror image expansion of the discontinuous virtual array to construct a three-dimensional discontinuous virtual cubic array and its corresponding virtual domain tensor, fully retaining the structured information of all discontinuous virtual domain statistics of the coprime array;
  • the present invention proposes a virtual domain tensor filling mechanism for non-continuous virtual arrays. By reconstructing the virtual domain tensor and dispersing its missing elements in pieces, the low-rank fillability of the virtual domain tensor is satisfied, thereby effectively filling the virtual domain tensor and realizing high-precision and high-resolution two-dimensional DOA estimation.
  • Fig. 1 is the overall flow chart of the present invention.
  • Fig. 2 is a schematic diagram of the coprime array structure constructed by the present invention.
  • Fig. 3 is a schematic diagram of the augmented discontinuous virtual array derived by the present invention.
  • Fig. 4 is a schematic diagram of a discontinuous virtual cubic array constructed by the present invention.
  • Fig. 5 is a performance comparison chart of direction of arrival estimation accuracy of the method proposed in the present invention under different signal-to-noise ratio conditions.
  • Fig. 6 is a comparison diagram of the direction of arrival estimation accuracy performance of the method proposed in the present invention under the condition of different sampling snapshot numbers.
  • the present invention proposes a two-dimensional direction of arrival estimation method for coprime arrays based on virtual domain tensor filling.
  • the two-dimensional direction of arrival estimation for coprime arrays based on Nyquist matching is realized.
  • the realization steps of the present invention are as follows:
  • Step 1 Construct a coprime surface array.
  • Step 2 Tensor modeling of received signal of coprime area array.
  • T sampling snapshot signals After superimposing the T sampling snapshot signals in the third dimension, a three-dimensional tensor signal can be obtained Modeled as:
  • s k [s k,1 ,s k,2 ,...,s k,T ] T is the multi-snapshot sampling signal waveform corresponding to the kth incident signal source
  • [ ] T represents the transpose operation
  • the vector outer product represents the vector outer product
  • the signal source of is expressed as:
  • the received signal can be three-dimensional tensor express:
  • Step 3 Construct the augmented discontinuous virtual array based on the cross-correlation tensor transformation of the coprime array. Due to the cross-correlation tensor Contains two sparse uniform sub-arrays corresponding to and The spatial information of In the dimension representing the spatial information of the same direction, the steering vectors corresponding to two sparse uniform sub-arrays can form a difference array on the exponent term, thereby constructing a two-dimensional augmented virtual area array.
  • the cross-correlation tensor The 1st and 3rd dimensions of (via the steering vector and Indicates) characterizes the spatial information in the x-axis direction, the 2nd and 4th dimensions (through the steering vector and Indicates) characterizes the spatial information in the y-axis direction; for this purpose, define a set of dimensions By pairing the cross-correlation tensor Perform tensor transformation of merging dimensions to obtain a virtual domain signal
  • non-continuous virtual array size is And contains holes in the entire row and column (ie: missing elements), as shown in Figure 3,
  • the cross-correlation noise tensor in about omitted in the theoretical modeling step of ; however, in practice, due to the use of the sampled cross-correlation tensor Alternative Theoretical Cross-Correlation Tensor Still covered in the statistical processing of virtual domain signals;
  • Step 4 Deduce the virtual domain tensor based on the mirror extension of the non-continuous virtual area array. Construct a non-continuous virtual area array About the Virtual Surface Array of Coordinate Axis Mirroring and will and Superimposed in the third dimension into a size of The three-dimensional discontinuous virtual cubic array of As shown in Figure 4. here, and Correspondingly, the virtual domain signal The conjugate transpose signal of The elements in are arranged to correspond to The position of the virtual array element in the middle can be obtained corresponding to the virtual array virtual domain signal Will and Superimpose on the third dimension to get the corresponding non-continuous virtual cubic array
  • Step 5 Scatter its missing elements into pieces by reconstructing the virtual domain tensor.
  • the virtual domain tensor In order to construct a virtual uniform cubic array for Nyquist-matched signal processing, the virtual domain tensor The slices of missing elements in are filled, thus corresponding to a virtual uniform cubic array
  • the low-rank tensor filling technology is based on the premise of random distribution of missing elements in the tensor, and cannot be used for virtual domain tensors with slices of missing elements. for efficient filling.
  • the value range of the translation window size is:
  • Subscripts that will have the same s y index subscript Superimpose in the fourth dimension to obtain L y four-dimensional tensors whose dimensions are P x ⁇ P y ⁇ 2 ⁇ L x ; further, superpose these L y four-dimensional tensors in the fifth dimension to obtain a five-dimensional virtual domain tensor
  • This five-dimensional virtual domain tensor Covers the spatial angle information of the x-axis and y-axis directions, the spatial mirror transformation information, and the spatial translation information of the x-axis and y-axis directions; the Merge along the 1st and 2nd dimensions of the representation space angle information, and merge along the 4th and 5th dimensions of the representation space translation information, and retain the 3rd dimension of the representation space mirror transformation information.
  • the specific operation is: define a set of dimensions then pass Dimensions are merged to obtain a three-dimensional reconstructed virtual domain tensor
  • the three dimensions of represent spatial angle information, spatial translation information, and spatial mirror transformation information respectively.
  • the virtual domain tensor Slices of missing elements in are randomly distributed to on the three spatial dimensions covered;
  • Step 6 Virtual Domain Tensor Filling Based on Tensor Kernel Norm Minimization.
  • the optimization variable is the padded virtual domain tensor corresponding to the virtual uniform cubic array
  • the tensor kernel norm express A collection of positional indices of non-missing elements in , Indicates tensors in on the mapping.
  • the kernel norm is a convex function
  • the virtual field tensor filling problem based on the minimization of the tensor kernel norm is a solvable convex optimization problem. Solving this convex optimization problem, we can get
  • Step 7 Decompose the filled virtual domain tensor to obtain the DOA estimation result.
  • padded virtual field tensor Can be expressed as:
  • ⁇ ( ⁇ ) represents the operation of taking the argument of a complex number
  • a (a) represents the ath element of a vector a; here, according to and The Kronecker structure of ⁇ 1 ⁇ [1,P x P y -1] and ⁇ 2 ⁇ [1,L x L y -1] respectively satisfy mod( ⁇ 1 ,P x ) ⁇ 0 and mod( ⁇ 2 ,P y ) ⁇ 0, and ⁇ 1 ⁇ [1,P x P y -P x ], ⁇ 2 ⁇ [1,L x L y -L x ], mod( ⁇ ) represents the remainder operation.
  • mod( ⁇ ) represents the remainder operation.
  • the closed-form solution of is:
  • the translation window size of the subtensor is 6 ⁇ 15 ⁇ 2. Assume that there are two narrow-band incident signals, and the azimuth and elevation angles of the incident direction are [30.6°, 25.6°] and [40.5°, 50.5°], respectively.
  • the method proposed in the present invention has performance advantages in the direction of arrival estimation accuracy no matter in different scenarios of expected signal-to-noise ratio (SNR) or in scenarios of different sampling snapshot numbers T.
  • the proposed method of the present invention makes full use of the structural information of the coprime area array receiving signal by constructing the virtual domain tensor, thereby having better DOA estimation performance;
  • the performance advantage of the proposed method of the present invention comes from filling and utilizing all non-continuous virtual domain statistical information through the virtual domain tensor, while the Tensor MUSIC method only extracts the continuous part of the non-continuous virtual array for virtual domain signal processing, resulting in the loss of virtual domain statistical information.
  • the present invention realizes the randomized distribution of missing elements in a piece through virtual domain tensor reconstruction, and based on this, designs a virtual domain tensor filling method based on the minimization of the tensor kernel norm, successfully utilizes all non-continuous virtual domain statistical information, and realizes high-precision two-dimensional DOA estimation of coprime arrays.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Variable-Direction Aerials And Aerial Arrays (AREA)

Abstract

一种基于虚拟域张量填充的互质面阵二维波达方向估计方法,主要解决现有方法中多维信号结构化信息丢失和虚拟域统计量无法完全利用的问题,其实现步骤是:构建互质面阵;互质面阵接收信号的张量建模;基于互质面阵互相关张量变换构造增广非连续虚拟面阵;基于非连续虚拟面阵的镜像拓展推导虚拟域张量;通过虚拟域张量重构分散其成片缺失元素;基于张量核范数最小化的虚拟域张量填充;对填充后的虚拟域张量进行分解获得波达方向估计结果。该方法充分利用互质面阵所对应的全部非连续虚拟域张量统计量信息,实现了高精度的波达方向估计,可用于目标定位。

Description

基于虚拟域张量填充的互质面阵二维波达方向估计方法 技术领域
本发明属于阵列信号处理技术领域,尤其涉及基于稀疏阵列虚拟域二阶高维统计量的统计信号处理技术,具体是一种基于虚拟域张量填充的互质面阵二维波达方向估计方法,可用于目标定位。
背景技术
互质阵列作为一种具备系统化结构的稀疏阵列,具有大孔径、高分辨率的优势,能够突破传统均匀阵列波达方向估计在估计性能与成本开销层面的性能瓶颈。由于互质阵列稀疏排布的阵元不满足奈奎斯特采样速率,为了实现奈奎斯特匹配的波达方向估计,常用做法是计算互质阵列接收信号的二阶统计量,以构造增广的非连续虚拟阵列,并从中提取连续部分,以实现基于虚拟域二阶等价信号的奈奎斯特匹配处理。进一步地,为了充分利用全部的非连续虚拟阵元,现有方法对非连续虚拟阵列进行填充,以实现波达方向估计性能提升。然而,上述方法通常将接收信号表示成矢量,并通过矢量化接收信号协方差矩阵推导虚拟域二阶等价信号;在部署互质面阵的场景中,由于互质面阵接收信号涵盖二维的空间信息,这种矢量化信号的处理方法破坏了互质面阵接收信号的原始空间信息结构,将会造成严重的性能损失。
为了保留多维接收信号的结构化信息,张量作为一种多维的数据类型,开始被应用于阵列信号处理领域,用于表征涵盖多维空间信息的接收信号,并对其进行特征分析和有效信息提取,从而实现高精度高分辨的波达方向估计。然而,当涉及到互质面阵虚拟域张量统计量处理,增广的多维非连续虚拟阵列将存在整片的孔洞,导致所对应的虚拟域张量存在成片缺失的元素。传统应用于图像修复的张量填充手段以张量中缺失元素的随机分布为前提,故无法有效填充虚拟域张量。为此,针对互质面阵的虚拟域张量模型,如何有效利用全部的非连续虚拟域张量统计量信息,实现高精度高分辨的二维波达方向估计,仍然 是一个亟待解决的问题。
发明内容
本发明的目的在于针对现有方法存在的多维信号结构化信息丢失和虚拟域统计量无法完全利用问题,提出一种基于虚拟域张量填充的互质面阵二维波达方向估计方法,为充分利用互质面阵所对应的全部非连续虚拟域张量统计量信息,以实现奈奎斯特匹配的高精度高分辨二维波达方向估计提供了可行的思路和有效的解决方案。
本发明的目的是通过以下技术方案来实现的:一种基于虚拟域张量填充的互质面阵二维波达方向估计方法,该方法包含以下步骤:
(1)接收端使用4M xM y+N xN y-1个物理天线阵元,按照互质面阵的结构进行架构;其中,M x、N x以及M y、N y分别为一对互质整数;该互质面阵分解为两个稀疏均匀子面阵
Figure PCTCN2022076430-appb-000001
Figure PCTCN2022076430-appb-000002
其中
Figure PCTCN2022076430-appb-000003
包含2M x×2M y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为N xd和N yd,
Figure PCTCN2022076430-appb-000004
包含N x×N y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为M xd和M yd,单位间隔d取为入射窄带信号波长λ的一半,即d=λ/2;
(2)假设有K个来自
Figure PCTCN2022076430-appb-000005
方向的远场窄带非相关信号源,θ k
Figure PCTCN2022076430-appb-000006
分别为第k个入射信号源的方位角和俯仰角,k=1,2,…,K,则稀疏均匀子面阵
Figure PCTCN2022076430-appb-000007
的T个采样快拍信号用一个三维张量
Figure PCTCN2022076430-appb-000008
表示为:
Figure PCTCN2022076430-appb-000009
其中,s k=[s k,1,s k,2,…,s k,T] T为对应第k个入射信号源的多快拍采样信号波形,[·] T表示转置操作,
Figure PCTCN2022076430-appb-000010
表示矢量外积,
Figure PCTCN2022076430-appb-000011
为与各信号源相互独立的噪声张量,
Figure PCTCN2022076430-appb-000012
Figure PCTCN2022076430-appb-000013
分别为
Figure PCTCN2022076430-appb-000014
在x轴和y轴方向上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000015
的信号源,表示为:
Figure PCTCN2022076430-appb-000016
Figure PCTCN2022076430-appb-000017
其中,
Figure PCTCN2022076430-appb-000018
Figure PCTCN2022076430-appb-000019
分别表示稀疏均匀子面阵
Figure PCTCN2022076430-appb-000020
在x轴和y轴方向上物理天线阵元的实际位置,且
Figure PCTCN2022076430-appb-000021
Figure PCTCN2022076430-appb-000022
稀疏均匀子面阵
Figure PCTCN2022076430-appb-000023
的T个采样快拍信号用另一个三维张量
Figure PCTCN2022076430-appb-000024
表示为:
Figure PCTCN2022076430-appb-000025
其中,
Figure PCTCN2022076430-appb-000026
为与各信号源相互独立的噪声张量,
Figure PCTCN2022076430-appb-000027
Figure PCTCN2022076430-appb-000028
分别为
Figure PCTCN2022076430-appb-000029
在x轴和y轴方向上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000030
的信号源,表示为:
Figure PCTCN2022076430-appb-000031
Figure PCTCN2022076430-appb-000032
其中,
Figure PCTCN2022076430-appb-000033
Figure PCTCN2022076430-appb-000034
分别表示稀疏均匀子面阵
Figure PCTCN2022076430-appb-000035
在x轴和y轴方向上物理天线阵元的实际位置,且
Figure PCTCN2022076430-appb-000036
通过求三维张量
Figure PCTCN2022076430-appb-000037
Figure PCTCN2022076430-appb-000038
的互相关统计量,得到二阶互相关张量
Figure PCTCN2022076430-appb-000039
Figure PCTCN2022076430-appb-000040
其中,
Figure PCTCN2022076430-appb-000041
表示第k个入射信号源的功率,
Figure PCTCN2022076430-appb-000042
表示互相关噪声张量,<·,·> r表示两个张量沿着第r维度的张量缩并操作,E[·]表 示取数学期望操作,(·) *表示共轭操作;互相关噪声张量
Figure PCTCN2022076430-appb-000043
仅在第(1,1,1,1)个位置上存在取值为
Figure PCTCN2022076430-appb-000044
的元素,
Figure PCTCN2022076430-appb-000045
表示噪声功率,而在其他位置上的元素取值均为0;
(3)定义维度集合
Figure PCTCN2022076430-appb-000046
通过对互相关张量
Figure PCTCN2022076430-appb-000047
进行维度合并的张量变换,得到一个虚拟域信号
Figure PCTCN2022076430-appb-000048
Figure PCTCN2022076430-appb-000049
其中,
Figure PCTCN2022076430-appb-000050
Figure PCTCN2022076430-appb-000051
Figure PCTCN2022076430-appb-000052
分别通过在指数项上形成差集数组,构造出沿着x轴和y轴方向的二维增广虚拟面阵,
Figure PCTCN2022076430-appb-000053
表示Kronecker积;因此,
Figure PCTCN2022076430-appb-000054
对应一个大小为
Figure PCTCN2022076430-appb-000055
的非连续虚拟面阵
Figure PCTCN2022076430-appb-000056
该非连续虚拟面阵
Figure PCTCN2022076430-appb-000057
中包含了整行和整列的孔洞,即缺失元素;
(4)构建非连续虚拟面阵
Figure PCTCN2022076430-appb-000058
关于坐标轴镜像的虚拟面阵
Figure PCTCN2022076430-appb-000059
并将
Figure PCTCN2022076430-appb-000060
Figure PCTCN2022076430-appb-000061
在第三维度上叠加成一个大小为
Figure PCTCN2022076430-appb-000062
的三维非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000063
这里,
Figure PCTCN2022076430-appb-000064
Figure PCTCN2022076430-appb-000065
对应地,将虚拟域信号
Figure PCTCN2022076430-appb-000066
的共轭转置信号
Figure PCTCN2022076430-appb-000067
中的元素进行重新排列,以对应
Figure PCTCN2022076430-appb-000068
中虚拟阵元的位置,得到对应于虚拟面阵
Figure PCTCN2022076430-appb-000069
的虚拟域信号
Figure PCTCN2022076430-appb-000070
Figure PCTCN2022076430-appb-000071
Figure PCTCN2022076430-appb-000072
在第三维度上进行叠加,得到对应非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000073
的虚拟域张量
Figure PCTCN2022076430-appb-000074
表示为:
Figure PCTCN2022076430-appb-000075
其中,
Figure PCTCN2022076430-appb-000076
Figure PCTCN2022076430-appb-000077
分别为非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000078
在x轴和y轴上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000079
的信号源,由于
Figure PCTCN2022076430-appb-000080
中存在孔洞,
Figure PCTCN2022076430-appb-000081
Figure PCTCN2022076430-appb-000082
分别对应
Figure PCTCN2022076430-appb-000083
中x轴和y轴方向上孔洞位置的元素置为零,
Figure PCTCN2022076430-appb-000084
表示对应于
Figure PCTCN2022076430-appb-000085
Figure PCTCN2022076430-appb-000086
的镜像变换因子矢量;由于非连续虚拟面阵
Figure PCTCN2022076430-appb-000087
中包含了整行 和整列的孔洞,由
Figure PCTCN2022076430-appb-000088
和其镜像部分
Figure PCTCN2022076430-appb-000089
叠加得到的非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000090
中包含了成片的孔洞,对应非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000091
的虚拟域张量
Figure PCTCN2022076430-appb-000092
由此包含了成片的缺失元素;
(5)设计一个大小为P x×P y×2的平移窗口选取虚拟域张量
Figure PCTCN2022076430-appb-000093
的一个子张量
Figure PCTCN2022076430-appb-000094
其中包含了
Figure PCTCN2022076430-appb-000095
三个维度上索引分别为(1:P x-1),(1:P y-1),(1:2)的元素;随后,将平移窗口分别沿x轴和y轴方向依次平移一个元素,将
Figure PCTCN2022076430-appb-000096
分割成L x×L y个子张量,表示为
Figure PCTCN2022076430-appb-000097
s x=1,2,…,L x,s y=1,2,…,L y;该平移窗口大小的取值范围为:
Figure PCTCN2022076430-appb-000098
Figure PCTCN2022076430-appb-000099
且L x、L y、P x、P y之间满足以下关系:
Figure PCTCN2022076430-appb-000100
Figure PCTCN2022076430-appb-000101
将具有相同s y索引下标的子张量
Figure PCTCN2022076430-appb-000102
在第四维度进行叠加,得到L y个维度为P x×P y×2×L x的四维张量;进一步地,将这L y个四维张量在第五维度叠加,得到一个五维虚拟域张量
Figure PCTCN2022076430-appb-000103
这个五维虚拟域张量
Figure PCTCN2022076430-appb-000104
涵盖了x轴和y轴方向空间角度信息、空间镜像变换信息,以及x轴和y轴方向的空间平移信息;定义维度集合
Figure PCTCN2022076430-appb-000105
则通过
Figure PCTCN2022076430-appb-000106
的维度合并,得到三维的重构虚拟域张量
Figure PCTCN2022076430-appb-000107
Figure PCTCN2022076430-appb-000108
Figure PCTCN2022076430-appb-000109
的三个维度分别表征空间角度信息、空间平移信息以及空间镜像变换信息,由此,原始虚拟域张量
Figure PCTCN2022076430-appb-000110
中的成片缺失元素被随机分布到
Figure PCTCN2022076430-appb-000111
所涵盖的三个空间维度上;
(6)设计一个基于张量核范数最小化的虚拟域张量填充优化问题:
Figure PCTCN2022076430-appb-000112
Figure PCTCN2022076430-appb-000113
其中,优化变量
Figure PCTCN2022076430-appb-000114
是填充后的虚拟域张量,对应于虚拟均匀立方阵列
Figure PCTCN2022076430-appb-000115
表示张量核范数,
Figure PCTCN2022076430-appb-000116
表示
Figure PCTCN2022076430-appb-000117
中非缺失元素的位置索引集合,
Figure PCTCN2022076430-appb-000118
表示张量在
Figure PCTCN2022076430-appb-000119
上的映射;由于核范数是凸函数,基于张量核范数最小化的虚拟域张量填充问题是一个可解的凸优化问题,求解该凸优化问题,得到
Figure PCTCN2022076430-appb-000120
(7)填充后的虚拟域张量
Figure PCTCN2022076430-appb-000121
表示为:
Figure PCTCN2022076430-appb-000122
其中,
Figure PCTCN2022076430-appb-000123
Figure PCTCN2022076430-appb-000124
的空间因子,
Figure PCTCN2022076430-appb-000125
Figure PCTCN2022076430-appb-000126
分别表示虚拟均匀立方阵列
Figure PCTCN2022076430-appb-000127
沿着x轴和y轴方向的导引矢量,
Figure PCTCN2022076430-appb-000128
Figure PCTCN2022076430-appb-000129
分别为平移窗口截取子张量过程中对应于x轴和y轴方向的空间平移因子矢量;对填充后的虚拟域张量
Figure PCTCN2022076430-appb-000130
进行canonical polyadic分解,得到三个因子矢量p k,q k和c k的估计值,表示为
Figure PCTCN2022076430-appb-000131
Figure PCTCN2022076430-appb-000132
从中提取包含在
Figure PCTCN2022076430-appb-000133
Figure PCTCN2022076430-appb-000134
指数项的角度参数,得到二维波达方向估计结果
Figure PCTCN2022076430-appb-000135
进一步地,步骤(1)所述的互质面阵结构具体描述为:在平面坐标系xoy上构造一对稀疏均匀子面阵
Figure PCTCN2022076430-appb-000136
Figure PCTCN2022076430-appb-000137
其中
Figure PCTCN2022076430-appb-000138
包含2M x×2M y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为N xd和N yd,其在xoy上的位置坐标为{(N xdm x,N ydm y),m x=0,1,...,2M x-1,m y=0,1,...,2M y-1};
Figure PCTCN2022076430-appb-000139
包含N x×N y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为M xd和M yd,其在xoy上的位置坐标为{(M xdn x,M ydn y),n x=0,1,...,N x-1,n y=0,1,...,N y-1};M x、N x以及M y、N y分别为一对互质整数;将
Figure PCTCN2022076430-appb-000140
Figure PCTCN2022076430-appb-000141
按照坐标系(0,0)位 置处阵元重叠的方式进行子阵列组合,获得实际包含4M xM y+N xN y-1个物理天线阵元的互质面阵。
进一步地,步骤(2)所述的互相关张量推导,在实际中,
Figure PCTCN2022076430-appb-000142
通过计算张量
Figure PCTCN2022076430-appb-000143
Figure PCTCN2022076430-appb-000144
的互相关统计量近似得到,即采样互相关张量
Figure PCTCN2022076430-appb-000145
Figure PCTCN2022076430-appb-000146
进一步地,步骤(7)中,对填充后的虚拟域张量
Figure PCTCN2022076430-appb-000147
进行canonical polyadic分解,得到因子矢量
Figure PCTCN2022076430-appb-000148
Figure PCTCN2022076430-appb-000149
则参数
Figure PCTCN2022076430-appb-000150
Figure PCTCN2022076430-appb-000151
Figure PCTCN2022076430-appb-000152
Figure PCTCN2022076430-appb-000153
中提取为:
Figure PCTCN2022076430-appb-000154
Figure PCTCN2022076430-appb-000155
其中,∠(·)表示复数取幅角操作,a (a)表示一个矢量a的第a个元素;这里,根据
Figure PCTCN2022076430-appb-000156
Figure PCTCN2022076430-appb-000157
的Kronecker结构,η 1∈[1,P xP y-1]和η 2∈[1,L xL y-1]分别满足mod(η 1,P x)≠0和mod(η 2,P y)≠0,且δ 1∈[1,P xP y-P x],δ 2∈[1,L xL y-L x],mod(·)表示取余数操作;根据参数(μ k,v k)与二维波达方向
Figure PCTCN2022076430-appb-000158
之间的关系,得到二维波达方向估计
Figure PCTCN2022076430-appb-000159
的闭式解为:
Figure PCTCN2022076430-appb-000160
Figure PCTCN2022076430-appb-000161
本发明与现有技术相比具有以下优点:
(1)本发明基于互相关张量推导增广的非连续虚拟面阵,并利用非连续虚拟面阵的镜像拓展构造三维非连续虚拟立方阵列及其对应的虚拟域张量,充分保留了互质阵列全部非连续虚拟域统计量的结构化信息;
(2)本发明提出了面向非连续虚拟阵列的虚拟域张量填充机制,通过虚拟域张量重构分散其成片缺失的元素,以满足虚拟域张量的低秩可填充性,从而对虚拟域张量进行有效填充,实现了高精度高分辨的二维波达方向估计。
附图说明
图1是本发明的总体流程框图。
图2是本发明所构建的互质面阵结构示意图。
图3是本发明所推导增广非连续虚拟面阵示意图。
图4是本发明所构造非连续虚拟立方阵列示意图。
图5是本发明所提方法在不同信噪比条件下的波达方向估计精度性能比较图。
图6是本发明所提方法在不同采样快拍数条件下的波达方向估计精度性能比较图。
具体实施方式
以下参照附图,对本发明的技术方案作进一步的详细说明。
为了解决现有方法存在的多维信号结构化信息丢失和虚拟域统计量无法完全利用问题,本发明提出了一种基于虚拟域张量填充的互质面阵二维波达方向估计方法,通过原始虚拟域张量成片缺失元素的有效填充,以实现奈奎斯特匹配的互质面阵二维波达方向估计。参照图1,本发明的实现步骤如下:
步骤1:构建互质面阵。在接收端使用4M xM y+N xN y-1个物理天线阵元构建互质面阵,如图2所示:在平面坐标系xoy上构造一对稀疏均匀子面阵
Figure PCTCN2022076430-appb-000162
Figure PCTCN2022076430-appb-000163
其中
Figure PCTCN2022076430-appb-000164
包含2M x×2M y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为N xd和N yd,其在xoy上的位置坐标为{(N xdm x,N ydm y),m x=0,1,...,2M x-1,m y=0,1,...,2M y-1};
Figure PCTCN2022076430-appb-000165
包含N x×N y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为M xd和M yd,其在xoy上的位置坐标为{(M xdn x,M ydn y),n x=0,1,...,N x-1,n y=0,1,...,N y-1};M x、N x以及M y、N y分别为一对互质整数;单位间隔d取为入射窄带信号波长λ的一半,即d=λ/2;将
Figure PCTCN2022076430-appb-000166
Figure PCTCN2022076430-appb-000167
按 照坐标系(0,0)位置处阵元重叠的方式进行子阵列组合,获得实际包含4M xM y+N xN y-1个物理天线阵元的互质面阵;
步骤2:互质面阵接收信号的张量建模。假设有K个来自
Figure PCTCN2022076430-appb-000168
Figure PCTCN2022076430-appb-000169
方向的远场窄带非相关信号源,θ k
Figure PCTCN2022076430-appb-000170
分别为第k个入射信号源的方位角和俯仰角,k=1,2,…,K,将互质面阵中稀疏均匀子面阵
Figure PCTCN2022076430-appb-000171
的T个采样快拍信号在第三个维度进行叠加后,可以得到一个三维张量信号
Figure PCTCN2022076430-appb-000172
Figure PCTCN2022076430-appb-000173
建模为:
Figure PCTCN2022076430-appb-000174
其中,s k=[s k,1,s k,2,…,s k,T] T为对应第k个入射信号源的多快拍采样信号波形,[·] T表示转置操作,
Figure PCTCN2022076430-appb-000175
表示矢量外积,
Figure PCTCN2022076430-appb-000176
为与各信号源相互独立的噪声张量,
Figure PCTCN2022076430-appb-000177
Figure PCTCN2022076430-appb-000178
分别为
Figure PCTCN2022076430-appb-000179
在x轴和y轴方向上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000180
的信号源,表示为:
Figure PCTCN2022076430-appb-000181
Figure PCTCN2022076430-appb-000182
其中,
Figure PCTCN2022076430-appb-000183
Figure PCTCN2022076430-appb-000184
分别表示稀疏均匀子面阵
Figure PCTCN2022076430-appb-000185
在x轴和y轴方向上物理天线阵元的实际位置,且
Figure PCTCN2022076430-appb-000186
Figure PCTCN2022076430-appb-000187
类似地,稀疏均匀子面阵
Figure PCTCN2022076430-appb-000188
的接收信号可用三维张量
Figure PCTCN2022076430-appb-000189
表示:
Figure PCTCN2022076430-appb-000190
其中,
Figure PCTCN2022076430-appb-000191
为与各信号源相互独立的噪声张量,
Figure PCTCN2022076430-appb-000192
Figure PCTCN2022076430-appb-000193
分别为
Figure PCTCN2022076430-appb-000194
在x轴和y轴方向上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000195
的信号源,表示为:
Figure PCTCN2022076430-appb-000196
Figure PCTCN2022076430-appb-000197
其中,
Figure PCTCN2022076430-appb-000198
Figure PCTCN2022076430-appb-000199
分别表示稀疏均匀子面阵
Figure PCTCN2022076430-appb-000200
在x轴和y轴方向上物理天线阵元的实际位置,且
Figure PCTCN2022076430-appb-000201
通过求三维张量信号
Figure PCTCN2022076430-appb-000202
Figure PCTCN2022076430-appb-000203
的互相关统计量,得到二阶互相关张量
Figure PCTCN2022076430-appb-000204
Figure PCTCN2022076430-appb-000205
其中,
Figure PCTCN2022076430-appb-000206
表示第k个入射信号源的功率,
Figure PCTCN2022076430-appb-000207
表示四维互相关噪声张量,<·,·> r表示两个张量沿着第r维度的张量缩并操作,E[·]表示取数学期望操作,(·) *表示共轭操作。这里,互相关噪声张量
Figure PCTCN2022076430-appb-000208
仅在第(1,1,1,1)个位置上存在取值为
Figure PCTCN2022076430-appb-000209
的元素,
Figure PCTCN2022076430-appb-000210
表示噪声功率,而在其他位置上的元素取值均为0。在实际中,
Figure PCTCN2022076430-appb-000211
通过计算张量信号
Figure PCTCN2022076430-appb-000212
Figure PCTCN2022076430-appb-000213
的互相关统计量近似得到,即采样互相关张量
Figure PCTCN2022076430-appb-000214
Figure PCTCN2022076430-appb-000215
步骤3:基于互质面阵的互相关张量变换构造增广非连续虚拟面阵。由于互相关张量
Figure PCTCN2022076430-appb-000216
中包含了对应两个稀疏均匀子面阵
Figure PCTCN2022076430-appb-000217
Figure PCTCN2022076430-appb-000218
的空间信息,通过合并
Figure PCTCN2022076430-appb-000219
中表征同一方向空间信息的维度,可以使对应两个稀疏均匀子面阵的导引矢量在指数项上形成差集数组,从而构造二维的增广虚拟面阵。具体地,互相关张量
Figure PCTCN2022076430-appb-000220
的第1、3维度(通过导引矢量
Figure PCTCN2022076430-appb-000221
Figure PCTCN2022076430-appb-000222
表示)表征x轴方向的空间信息,第2、4维度(通过导引矢量
Figure PCTCN2022076430-appb-000223
Figure PCTCN2022076430-appb-000224
表示)表征y轴方向的空间信息;为此,定义维度集合
Figure PCTCN2022076430-appb-000225
Figure PCTCN2022076430-appb-000226
通过对互相关张量
Figure PCTCN2022076430-appb-000227
进行维度合并的张量变换,得到一个虚拟域信号
Figure PCTCN2022076430-appb-000228
Figure PCTCN2022076430-appb-000229
其中,
Figure PCTCN2022076430-appb-000230
Figure PCTCN2022076430-appb-000231
Figure PCTCN2022076430-appb-000232
等价为非连续虚拟面阵
Figure PCTCN2022076430-appb-000233
在x轴和y轴上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000234
的信号源,
Figure PCTCN2022076430-appb-000235
表示Kronecker积。非连续虚拟面阵
Figure PCTCN2022076430-appb-000236
大小为
Figure PCTCN2022076430-appb-000237
且包含了整行和整列的孔洞(即:缺失元素),如图3所示,
Figure PCTCN2022076430-appb-000238
这里,为了简化推导过程,互相关噪声张量
Figure PCTCN2022076430-appb-000239
在关于
Figure PCTCN2022076430-appb-000240
的理论建模步骤中省略;然而,在实际中,由于使用采样互相关张量
Figure PCTCN2022076430-appb-000241
替代理论互相关张量
Figure PCTCN2022076430-appb-000242
仍旧涵盖于虚拟域信号统计处理过程中;
步骤4:基于非连续虚拟面阵的镜像拓展推导虚拟域张量。构建非连续虚拟面阵
Figure PCTCN2022076430-appb-000243
关于坐标轴镜像的虚拟面阵
Figure PCTCN2022076430-appb-000244
并将
Figure PCTCN2022076430-appb-000245
Figure PCTCN2022076430-appb-000246
在第三维度上叠加成一个大小为
Figure PCTCN2022076430-appb-000247
的三维非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000248
如图4所示。这里,
Figure PCTCN2022076430-appb-000249
Figure PCTCN2022076430-appb-000250
Figure PCTCN2022076430-appb-000251
对应地,将虚拟域信号
Figure PCTCN2022076430-appb-000252
的共轭转置信号
Figure PCTCN2022076430-appb-000253
中的元素进行排列,以对应
Figure PCTCN2022076430-appb-000254
中虚拟阵元的位置,即可得到对应于虚拟面阵
Figure PCTCN2022076430-appb-000255
的虚拟域信号
Figure PCTCN2022076430-appb-000256
Figure PCTCN2022076430-appb-000257
Figure PCTCN2022076430-appb-000258
在第三维度上进行叠加,得到对应非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000259
的虚拟域张量
Figure PCTCN2022076430-appb-000260
表示为:
Figure PCTCN2022076430-appb-000261
其中,
Figure PCTCN2022076430-appb-000262
Figure PCTCN2022076430-appb-000263
分别为非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000264
在x轴和y轴上的导引矢量,对应于来波方向为
Figure PCTCN2022076430-appb-000265
的信号源;由于
Figure PCTCN2022076430-appb-000266
中存在缺失元素(孔洞),
Figure PCTCN2022076430-appb-000267
Figure PCTCN2022076430-appb-000268
分别对应
Figure PCTCN2022076430-appb-000269
中x轴和y轴方向上孔洞位置的元素置为零,
Figure PCTCN2022076430-appb-000270
表示对应
Figure PCTCN2022076430-appb-000271
Figure PCTCN2022076430-appb-000272
的镜像变换因子矢量;由于非连续虚拟面阵
Figure PCTCN2022076430-appb-000273
中包含了整行和整列的孔洞,由
Figure PCTCN2022076430-appb-000274
和其镜像部分
Figure PCTCN2022076430-appb-000275
叠加得到的非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000276
中包含了成片的孔洞,对应非连续虚拟立方阵列
Figure PCTCN2022076430-appb-000277
的虚拟域张量
Figure PCTCN2022076430-appb-000278
由此包含了成片的缺失元素;
步骤5:通过虚拟域张量重构分散其成片缺失元素。为了构造一个虚拟均匀立方阵列以实现奈奎斯特匹配的信号处理,需要对虚拟域张量
Figure PCTCN2022076430-appb-000279
中的成片缺失元素进行填充,从而对应一个虚拟均匀立方阵列
Figure PCTCN2022076430-appb-000280
然而,低秩张量填充技术以张量中缺失元素随机化分布为前提,无法对存在成片缺失元素的虚拟域张量
Figure PCTCN2022076430-appb-000281
进行有效填充。为此,通过重构虚拟域张量
Figure PCTCN2022076430-appb-000282
分散其成片缺失元素,具体过程为:通过设计一个大小为P x×P y×2的平移窗口选取虚拟域张量
Figure PCTCN2022076430-appb-000283
的一个子张量
Figure PCTCN2022076430-appb-000284
其中包含了
Figure PCTCN2022076430-appb-000285
三个维度上索引分别为(1:P x-1),(1:P y-1),(1:2)的元素;随后,将平移窗口分别沿x轴和y轴方向依次平移一个元素,则可以将
Figure PCTCN2022076430-appb-000286
分割成L x×L y个子张量,表示为
Figure PCTCN2022076430-appb-000287
s y=1,2,…,L y。该平移窗口大小的取值范围为:
Figure PCTCN2022076430-appb-000288
Figure PCTCN2022076430-appb-000289
且L x、L y、P x、P y之间满足以下关系:
Figure PCTCN2022076430-appb-000290
Figure PCTCN2022076430-appb-000291
将具有相同s y索引下标的子张量
Figure PCTCN2022076430-appb-000292
在第四维度进行叠加,得到L y个维度为P x×P y×2×L x的四维张量;进一步地,将这L y个四维张量在第五维度叠加,得到一个五维虚拟域张量
Figure PCTCN2022076430-appb-000293
这个五维虚拟域张量
Figure PCTCN2022076430-appb-000294
涵盖了x轴和y轴方向空间角度信息、空间镜像变换信息,以及x轴和y轴方向的空间平移信息;将
Figure PCTCN2022076430-appb-000295
沿着表征空间角度信息的第1、2维度进行合并,同时沿着表征空间平移信息的第4、5维度进行合并,并保留表征空间镜像变换信息的第3维度,具体操作为:定义维度集合
Figure PCTCN2022076430-appb-000296
则通过
Figure PCTCN2022076430-appb-000297
的维 度合并,可得到三维的重构虚拟域张量
Figure PCTCN2022076430-appb-000298
Figure PCTCN2022076430-appb-000299
Figure PCTCN2022076430-appb-000300
的三个维度分别表征空间角度信息、空间平移信息以及空间镜像变换信息,由此,虚拟域张量
Figure PCTCN2022076430-appb-000301
中的成片缺失元素被随机分布到
Figure PCTCN2022076430-appb-000302
所涵盖的三个空间维度上;
步骤6:基于张量核范数最小化的虚拟域张量填充。为了对重构的虚拟域张量
Figure PCTCN2022076430-appb-000303
进行填充,设计如下基于张量核范数最小化的虚拟域张量填充优化问题:
Figure PCTCN2022076430-appb-000304
Figure PCTCN2022076430-appb-000305
其中,优化变量
Figure PCTCN2022076430-appb-000306
是填充后的虚拟域张量,对应于虚拟均匀立方阵列
Figure PCTCN2022076430-appb-000307
表示张量核范数,
Figure PCTCN2022076430-appb-000308
表示
Figure PCTCN2022076430-appb-000309
中非缺失元素的位置索引集合,
Figure PCTCN2022076430-appb-000310
表示张量在
Figure PCTCN2022076430-appb-000311
上的映射。由于核范数是凸函数,该基于张量核范数最小化的虚拟域张量填充问题是一个可解的凸优化问题。求解该凸优化问题,即可得到
Figure PCTCN2022076430-appb-000312
步骤7:对填充后的虚拟域张量进行分解获得波达方向估计结果。填充后的虚拟域张量
Figure PCTCN2022076430-appb-000313
可表示为:
Figure PCTCN2022076430-appb-000314
其中,
Figure PCTCN2022076430-appb-000315
Figure PCTCN2022076430-appb-000316
的空间因子,
Figure PCTCN2022076430-appb-000317
Figure PCTCN2022076430-appb-000318
分别表示虚拟均匀立方阵列
Figure PCTCN2022076430-appb-000319
沿着x轴和y轴方向的导引矢量,
Figure PCTCN2022076430-appb-000320
Figure PCTCN2022076430-appb-000321
分别为平移窗口截取子张量过程中对应于x轴和y轴方向的空间平移因子矢量。对填充后的虚拟域张量
Figure PCTCN2022076430-appb-000322
进行canonical polyadic分解,即可得到三个因子矢量p k,q k和c k的估计值,表示为
Figure PCTCN2022076430-appb-000323
Figure PCTCN2022076430-appb-000324
则参数
Figure PCTCN2022076430-appb-000325
Figure PCTCN2022076430-appb-000326
可从
Figure PCTCN2022076430-appb-000327
Figure PCTCN2022076430-appb-000328
中提取得到:
Figure PCTCN2022076430-appb-000329
Figure PCTCN2022076430-appb-000330
其中,∠(·)表示复数取幅角操作,a (a)表示一个矢量a的第a个元素;这里,根据
Figure PCTCN2022076430-appb-000331
Figure PCTCN2022076430-appb-000332
的Kronecker结构,η 1∈[1,P xP y-1]和η 2∈[1,L xL y-1]分别满足mod(η 1,P x)≠0和mod(η 2,P y)≠0,且δ 1∈[1,P xP y-P x],δ 2∈[1,L xL y-L x],mod(·)表示取余数操作。根据参数(μ k,v k)与二维波达方向
Figure PCTCN2022076430-appb-000333
之间的关系,得到二维波达方向估计
Figure PCTCN2022076430-appb-000334
的闭式解为:
Figure PCTCN2022076430-appb-000335
Figure PCTCN2022076430-appb-000336
下面结合仿真实例对本发明的效果做进一步的描述。
仿真实例:采用互质面阵接收入射信号,其参数选取为M x=2,M y=3,N x=3,N y=4,即架构的互质面阵共包含4M xM y+N xN y-1=35个物理阵元。子张量的平移窗口大小为6×15×2。假定有2个窄带入射信号,入射方向的方位角和俯仰角分别是[30.6°,25.6°]和[40.5°,50.5°]。将本发明所提基于虚拟域张量填充的互质面阵二维波达方向估计方法与传统仅利用虚拟域连续部分的多重信号分类(Multiple Signal Classification,MUSIC)方法和张量多重信号分类(Tensor MUSIC)方法进行对比,在采样快拍数T=300条件下,绘制均方根误差(Root-mean-square Error,RMSE)随信噪比SNR变化的性能对比曲线,如图5所示;在SNR=0dB条件下,绘制RMSE随采样快拍数T变化的性能对比曲线,如图6所示。
从图5和图6的对比结果可以看出,无论是在不同的期望信号信噪比SNR场景,还是在不同的采样快拍数T场景下,本发明所提方法在波达方向估计精度 上均存在性能优势。相比于传统的MUSIC方法,本发明所提方法通过构建虚拟域张量,充分利用了互质面阵接收信号的结构化信息,从而具备更优的波达方向估计性能;相比于Tensor MUSIC方法,本发明所提方法的性能优势来源于通过虚拟域张量填充利用全部的非连续虚拟域统计量信息,而Tensor MUSIC方法只提取非连续虚拟阵列的连续部分进行虚拟域信号处理,造成了虚拟域统计量信息的丢失。
综上所述,本发明通过虚拟域张量重构实现成片缺失元素随机化分布,并以此为基础,设计基于张量核范数最小化的虚拟域张量填充方法,成功利用了全部的非连续虚拟域统计量信息,实现了高精度的互质面阵二维波达方向估计。
以上所述仅是本发明的优选实施方式,虽然本发明已以较佳实施例披露如上,然而并非用以限定本发明。任何熟悉本领域的技术人员,在不脱离本发明技术方案范围情况下,都可利用上述揭示的方法和技术内容对本发明技术方案做出许多可能的变动和修饰,或修改为等同变化的等效实施例。因此,凡是未脱离本发明技术方案的内容,依据本发明的技术实质对以上实施例所做的任何的简单修改、等同变化及修饰,均仍属于本发明技术方案保护的范围内。

Claims (4)

  1. 一种基于虚拟域张量填充的互质面阵二维波达方向估计方法,其特征在于,包含以下步骤:
    (1)接收端使用4M xM y+N xN y-1个物理天线阵元,按照互质面阵的结构进行架构;其中,M x、N x以及M y、N y分别为一对互质整数;该互质面阵分解为两个稀疏均匀子面阵
    Figure PCTCN2022076430-appb-100001
    Figure PCTCN2022076430-appb-100002
    其中
    Figure PCTCN2022076430-appb-100003
    包含2M x×2M y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为N xd和N yd,
    Figure PCTCN2022076430-appb-100004
    包含N x×N y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为M xd和M yd,单位间隔d取为入射窄带信号波长λ的一半,即d=λ/2;
    (2)假设有K个来自
    Figure PCTCN2022076430-appb-100005
    方向的远场窄带非相关信号源,θ k
    Figure PCTCN2022076430-appb-100006
    分别为第k个入射信号源的方位角和俯仰角,k=1,2,…,K,则稀疏均匀子面阵
    Figure PCTCN2022076430-appb-100007
    的T个采样快拍信号用一个三维张量
    Figure PCTCN2022076430-appb-100008
    表示为:
    Figure PCTCN2022076430-appb-100009
    其中,s k=[s k,1,s k,2,…,s k,T] T为对应第k个入射信号源的多快拍采样信号波形,[·] T表示转置操作,°表示矢量外积,
    Figure PCTCN2022076430-appb-100010
    为与各信号源相互独立的噪声张量,
    Figure PCTCN2022076430-appb-100011
    Figure PCTCN2022076430-appb-100012
    分别为
    Figure PCTCN2022076430-appb-100013
    在x轴和y轴方向上的导引矢量,对应于来波方向为
    Figure PCTCN2022076430-appb-100014
    的信号源,表示为:
    Figure PCTCN2022076430-appb-100015
    Figure PCTCN2022076430-appb-100016
    其中,
    Figure PCTCN2022076430-appb-100017
    Figure PCTCN2022076430-appb-100018
    分别表示稀疏均匀子面阵
    Figure PCTCN2022076430-appb-100019
    在x轴和y轴方向上物理天线阵元的实际位置,且
    Figure PCTCN2022076430-appb-100020
    Figure PCTCN2022076430-appb-100021
    稀疏均匀子面阵
    Figure PCTCN2022076430-appb-100022
    的T个采样快拍信号用另一个三维张量
    Figure PCTCN2022076430-appb-100023
    表示为:
    Figure PCTCN2022076430-appb-100024
    其中,
    Figure PCTCN2022076430-appb-100025
    为与各信号源相互独立的噪声张量,
    Figure PCTCN2022076430-appb-100026
    Figure PCTCN2022076430-appb-100027
    分别为
    Figure PCTCN2022076430-appb-100028
    在x轴和y轴方向上的导引矢量,对应于来波方向为
    Figure PCTCN2022076430-appb-100029
    的信号源,表示为:
    Figure PCTCN2022076430-appb-100030
    Figure PCTCN2022076430-appb-100031
    其中,
    Figure PCTCN2022076430-appb-100032
    Figure PCTCN2022076430-appb-100033
    分别表示稀疏均匀子面阵
    Figure PCTCN2022076430-appb-100034
    在x轴和y轴方向上物理天线阵元的实际位置,且
    Figure PCTCN2022076430-appb-100035
    通过求三维张量
    Figure PCTCN2022076430-appb-100036
    Figure PCTCN2022076430-appb-100037
    的互相关统计量,得到二阶互相关张量
    Figure PCTCN2022076430-appb-100038
    Figure PCTCN2022076430-appb-100039
    其中,
    Figure PCTCN2022076430-appb-100040
    表示第k个入射信号源的功率,
    Figure PCTCN2022076430-appb-100041
    表示互相关噪声张量,<·,·> r表示两个张量沿着第r维度的张量缩并操作,E[·]表示取数学期望操作,(·) *表示共轭操作;互相关噪声张量
    Figure PCTCN2022076430-appb-100042
    仅在第(1,1,1,1)个位置上存在取值为
    Figure PCTCN2022076430-appb-100043
    的元素,
    Figure PCTCN2022076430-appb-100044
    表示噪声功率,而在其他位置上的元素取值均为0;
    (3)定义维度集合
    Figure PCTCN2022076430-appb-100045
    通过对互相关张量
    Figure PCTCN2022076430-appb-100046
    进行维度合并的张量变换,得到一个虚拟域信号
    Figure PCTCN2022076430-appb-100047
    Figure PCTCN2022076430-appb-100048
    其中,
    Figure PCTCN2022076430-appb-100049
    Figure PCTCN2022076430-appb-100050
    Figure PCTCN2022076430-appb-100051
    分别通过在指数项上形成差集数组,构造出沿着x轴和y轴方向的二维增广虚拟面阵,
    Figure PCTCN2022076430-appb-100052
    表示Kronecker积;因此,
    Figure PCTCN2022076430-appb-100053
    对应一个大小为
    Figure PCTCN2022076430-appb-100054
    的非连续虚拟面阵
    Figure PCTCN2022076430-appb-100055
    该非连续虚拟面阵
    Figure PCTCN2022076430-appb-100056
    中包含了整行和整列的孔洞,即缺失元素;
    (4)构建非连续虚拟面阵
    Figure PCTCN2022076430-appb-100057
    关于坐标轴镜像的虚拟面阵
    Figure PCTCN2022076430-appb-100058
    并将
    Figure PCTCN2022076430-appb-100059
    Figure PCTCN2022076430-appb-100060
    在第三维度上叠加成一个大小为
    Figure PCTCN2022076430-appb-100061
    的三维非连续虚拟立方阵列
    Figure PCTCN2022076430-appb-100062
    这里,
    Figure PCTCN2022076430-appb-100063
    Figure PCTCN2022076430-appb-100064
    对应地,将虚拟域信号
    Figure PCTCN2022076430-appb-100065
    的共轭转置信号
    Figure PCTCN2022076430-appb-100066
    中的元素进行重新排列,以对应
    Figure PCTCN2022076430-appb-100067
    中虚拟阵元的位置,得到对应于虚拟面阵
    Figure PCTCN2022076430-appb-100068
    的虚拟域信号
    Figure PCTCN2022076430-appb-100069
    Figure PCTCN2022076430-appb-100070
    Figure PCTCN2022076430-appb-100071
    在第三维度上进行叠加,得到对应非连续虚拟立方阵列
    Figure PCTCN2022076430-appb-100072
    的虚拟域张量
    Figure PCTCN2022076430-appb-100073
    表示为:
    Figure PCTCN2022076430-appb-100074
    其中,
    Figure PCTCN2022076430-appb-100075
    Figure PCTCN2022076430-appb-100076
    分别为非连续虚拟立方阵列
    Figure PCTCN2022076430-appb-100077
    在x轴和y轴上的导引矢量,对应于来波方向为
    Figure PCTCN2022076430-appb-100078
    的信号源,由于
    Figure PCTCN2022076430-appb-100079
    中存在孔洞,
    Figure PCTCN2022076430-appb-100080
    Figure PCTCN2022076430-appb-100081
    分别对应
    Figure PCTCN2022076430-appb-100082
    中x轴和y轴方向上孔洞位置的元素置为零,
    Figure PCTCN2022076430-appb-100083
    表示对应于
    Figure PCTCN2022076430-appb-100084
    Figure PCTCN2022076430-appb-100085
    的镜像变换因子矢量;由于非连续虚拟面阵
    Figure PCTCN2022076430-appb-100086
    中包含了整行和整列的孔洞,由
    Figure PCTCN2022076430-appb-100087
    和其镜像部分
    Figure PCTCN2022076430-appb-100088
    叠加得到的非连续虚拟立方阵列
    Figure PCTCN2022076430-appb-100089
    中包含了成片的孔洞,对应非连续虚拟立方阵列
    Figure PCTCN2022076430-appb-100090
    的虚拟域张量
    Figure PCTCN2022076430-appb-100091
    由此包含了成片的缺失元素;
    (5)设计一个大小为P x×P y×2的平移窗口选取虚拟域张量
    Figure PCTCN2022076430-appb-100092
    的一个子张量
    Figure PCTCN2022076430-appb-100093
    中包含了
    Figure PCTCN2022076430-appb-100094
    三个维度上索引分别为(1:P x-1),(1:P y-1),(1:2) 的元素;随后,将平移窗口分别沿x轴和y轴方向依次平移一个元素,将
    Figure PCTCN2022076430-appb-100095
    分割成L x×L y个子张量,表示为
    Figure PCTCN2022076430-appb-100096
    该平移窗口大小的取值范围为:
    Figure PCTCN2022076430-appb-100097
    Figure PCTCN2022076430-appb-100098
    且L x、L y、P x、P y之间满足以下关系:
    Figure PCTCN2022076430-appb-100099
    Figure PCTCN2022076430-appb-100100
    将具有相同s y索引下标的子张量
    Figure PCTCN2022076430-appb-100101
    在第四维度进行叠加,得到L y个维度为P x×P y×2×L x的四维张量;进一步地,将这L y个四维张量在第五维度叠加,得到一个五维虚拟域张量
    Figure PCTCN2022076430-appb-100102
    这个五维虚拟域张量
    Figure PCTCN2022076430-appb-100103
    涵盖了x轴和y轴方向空间角度信息、空间镜像变换信息,以及x轴和y轴方向的空间平移信息;定义维度集合
    Figure PCTCN2022076430-appb-100104
    则通过
    Figure PCTCN2022076430-appb-100105
    的维度合并,得到三维的重构虚拟域张量
    Figure PCTCN2022076430-appb-100106
    Figure PCTCN2022076430-appb-100107
    Figure PCTCN2022076430-appb-100108
    的三个维度分别表征空间角度信息、空间平移信息以及空间镜像变换信息,由此,原始虚拟域张量
    Figure PCTCN2022076430-appb-100109
    中的成片缺失元素被随机分布到
    Figure PCTCN2022076430-appb-100110
    所涵盖的三个空间维度上;
    (6)设计一个基于张量核范数最小化的虚拟域张量填充优化问题:
    Figure PCTCN2022076430-appb-100111
    Figure PCTCN2022076430-appb-100112
    其中,优化变量
    Figure PCTCN2022076430-appb-100113
    是填充后的虚拟域张量,对应于虚拟均匀立方阵列
    Figure PCTCN2022076430-appb-100114
    表示张量核范数,
    Figure PCTCN2022076430-appb-100115
    表示
    Figure PCTCN2022076430-appb-100116
    中非缺失元素的位置索引集合,
    Figure PCTCN2022076430-appb-100117
    表示张量在
    Figure PCTCN2022076430-appb-100118
    上的映射;由于核范数是凸函数,基于张量核范数最小化的虚拟域张量填充问题是一个可解的凸优化问题,求解该凸优化问题,得到
    Figure PCTCN2022076430-appb-100119
    (7)填充后的虚拟域张量
    Figure PCTCN2022076430-appb-100120
    表示为:
    Figure PCTCN2022076430-appb-100121
    其中,
    Figure PCTCN2022076430-appb-100122
    Figure PCTCN2022076430-appb-100123
    的空间因子,
    Figure PCTCN2022076430-appb-100124
    Figure PCTCN2022076430-appb-100125
    分别表示虚拟均匀立方阵列
    Figure PCTCN2022076430-appb-100126
    沿着x轴和y轴方向的导引矢量,
    Figure PCTCN2022076430-appb-100127
    Figure PCTCN2022076430-appb-100128
    分别为平移窗口截取子张量过程中对应于x轴和y轴方向的空间平移因子矢量;对填充后的虚拟域张量
    Figure PCTCN2022076430-appb-100129
    进行canonical polyadic分解,得到三个因子矢量p k,q k和c k的估计值,表示为
    Figure PCTCN2022076430-appb-100130
    Figure PCTCN2022076430-appb-100131
    从中提取包含在
    Figure PCTCN2022076430-appb-100132
    Figure PCTCN2022076430-appb-100133
    指数项的角度参数,得到二维波达方向估计结果
    Figure PCTCN2022076430-appb-100134
  2. 根据权利要求1所述的基于虚拟域张量填充的互质面阵二维波达方向估计方法,其特征在于,步骤(1)所述的互质面阵结构具体描述为:在平面坐标系xoy上构造一对稀疏均匀子面阵
    Figure PCTCN2022076430-appb-100135
    Figure PCTCN2022076430-appb-100136
    其中
    Figure PCTCN2022076430-appb-100137
    包含2M x×2M y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为N xd和N yd,其在xoy上的位置坐标为{(N xdm x,N ydm y),m x=0,1,...,2M x-1,m y=0,1,...,2M y-1};
    Figure PCTCN2022076430-appb-100138
    包含N x×N y个天线阵元,在x轴方向上和y轴方向上的阵元间距分别为M xd和M yd,其在xoy上的位置坐标为{(M xdn x,M ydn y),n x=0,1,...,N x-1,n y=0,1,...,N y-1};M x、N x以及M y、N y分别为一对互质整数;将
    Figure PCTCN2022076430-appb-100139
    Figure PCTCN2022076430-appb-100140
    按照坐标系(0,0)位置处阵元重叠的方式进行子阵列组合,获得实际包含4M xM y+N xN y-1个物理天线阵元的互质面阵。
  3. 根据权利要求1所述的基于虚拟域张量填充的互质面阵二维波达方向估计方法,其特征在于,步骤(2)所述的互相关张量推导,在实际中,
    Figure PCTCN2022076430-appb-100141
    通过计算张量
    Figure PCTCN2022076430-appb-100142
    Figure PCTCN2022076430-appb-100143
    的互相关统计量近似得到,即采样互相关张量
    Figure PCTCN2022076430-appb-100144
    Figure PCTCN2022076430-appb-100145
    Figure PCTCN2022076430-appb-100146
  4. 根据权利要求1所述的基于虚拟域张量填充的互质面阵二维波达方向估计方法,其特征在于,步骤(7)中,对填充后的虚拟域张量
    Figure PCTCN2022076430-appb-100147
    进行canonical polyadic分解,得到因子矢量
    Figure PCTCN2022076430-appb-100148
    Figure PCTCN2022076430-appb-100149
    则参数
    Figure PCTCN2022076430-appb-100150
    Figure PCTCN2022076430-appb-100151
    Figure PCTCN2022076430-appb-100152
    Figure PCTCN2022076430-appb-100153
    中提取为:
    Figure PCTCN2022076430-appb-100154
    Figure PCTCN2022076430-appb-100155
    其中,∠(·)表示复数取幅角操作,a (a)表示一个矢量a的第a个元素;这里,根据
    Figure PCTCN2022076430-appb-100156
    Figure PCTCN2022076430-appb-100157
    的Kronecker结构,η 1∈[1,P xP y-1]和η 2∈[1,L xL y-1]分别满足mod(η 1,P x)≠0和mod(η 2,P y)≠0,且δ 1∈[1,P xP y-P x],δ 2∈[1,L xL y-L x],mod(·)表示取余数操作;根据参数(μ k,v k)与二维波达方向
    Figure PCTCN2022076430-appb-100158
    之间的关系,得到二维波达方向估计
    Figure PCTCN2022076430-appb-100159
    的闭式解为:
    Figure PCTCN2022076430-appb-100160
    Figure PCTCN2022076430-appb-100161
PCT/CN2022/076430 2022-01-21 2022-02-16 基于虚拟域张量填充的互质面阵二维波达方向估计方法 WO2023137812A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/920,401 US20240210510A1 (en) 2022-01-21 2022-02-16 Two-dimensional direction-of-arrival estimation method for coprime surface array based on virtual domain tensor filling

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210077881.1A CN114444298A (zh) 2022-01-21 2022-01-21 基于虚拟域张量填充的互质面阵二维波达方向估计方法
CN202210077881.1 2022-01-21

Publications (1)

Publication Number Publication Date
WO2023137812A1 true WO2023137812A1 (zh) 2023-07-27

Family

ID=81369705

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/076430 WO2023137812A1 (zh) 2022-01-21 2022-02-16 基于虚拟域张量填充的互质面阵二维波达方向估计方法

Country Status (3)

Country Link
US (1) US20240210510A1 (zh)
CN (1) CN114444298A (zh)
WO (1) WO2023137812A1 (zh)

Citations (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030058153A1 (en) * 2001-09-07 2003-03-27 Lockheed Martin Corporation Adaptive digital beamforming radar method and system for maintaining multiple source angle super-resolution capability in jamming
CN104749552A (zh) * 2015-03-21 2015-07-01 西安电子科技大学 基于稀疏重构的互质阵列波达方向角估计方法
US20160172767A1 (en) * 2014-12-12 2016-06-16 The Boeing Company Congruent non-uniform antenna arrays
CN106896340A (zh) * 2017-01-20 2017-06-27 浙江大学 一种基于压缩感知的互质阵列高精度波达方向估计方法
CN107037392A (zh) * 2017-03-01 2017-08-11 浙江大学 一种基于压缩感知的自由度增加型互质阵列波达方向估计方法
CN107102291A (zh) * 2017-05-03 2017-08-29 浙江大学 基于虚拟阵列内插的无网格化互质阵列波达方向估计方法
CN107315160A (zh) * 2017-05-03 2017-11-03 浙江大学 基于内插虚拟阵列信号原子范数最小化的互质阵列波达方向估计方法
CN107329108A (zh) * 2017-05-03 2017-11-07 浙江大学 基于内插虚拟阵列协方差矩阵Toeplitz化重建的互质阵列波达方向估计方法
CN107422295A (zh) * 2017-08-30 2017-12-01 浙江大学 基于互质阵列虚拟域等价信号原子范数表示的波达方向估计方法
CN107589399A (zh) * 2017-08-24 2018-01-16 浙江大学 基于多采样虚拟信号奇异值分解的互质阵列波达方向估计方法
CN108344967A (zh) * 2018-01-20 2018-07-31 中国人民解放军战略支援部队信息工程大学 基于互质面阵的二维波达方向快速估计方法
CN108710102A (zh) * 2018-05-15 2018-10-26 浙江大学 基于互质阵列二阶等价虚拟信号离散傅里叶逆变换的波达方向估计方法
CN109143152A (zh) * 2018-09-25 2019-01-04 哈尔滨工业大学 基于张量建模的极化阵列波达方向和极化参数估计方法
CN109143228A (zh) * 2018-08-28 2019-01-04 海南大学 互耦条件下双基地mimo雷达非圆目标的角度估计方法
CN110244259A (zh) * 2019-06-26 2019-09-17 南京航空航天大学 数据缺失情况下基于低n-秩最小化张量填充的二维到达角估计方法
CN110927661A (zh) * 2019-11-22 2020-03-27 重庆邮电大学 基于music算法的单基地展开互质阵列mimo雷达doa估计方法
CN111610486A (zh) * 2020-05-03 2020-09-01 浙江大学 基于平面互质阵列虚拟域张量空间谱搜索的高分辨精确二维波达方向估计方法
CN111624545A (zh) * 2020-05-03 2020-09-04 浙江大学 基于结构化虚拟域张量信号处理的互质面阵二维波达方向估计方法
CN112904272A (zh) * 2021-01-18 2021-06-04 浙江大学 基于互相关张量的三维互质立方阵列波达方向估计方法
CN113552532A (zh) * 2021-07-09 2021-10-26 浙江大学 基于耦合张量分解的l型互质阵列波达方向估计方法

Patent Citations (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030058153A1 (en) * 2001-09-07 2003-03-27 Lockheed Martin Corporation Adaptive digital beamforming radar method and system for maintaining multiple source angle super-resolution capability in jamming
US20160172767A1 (en) * 2014-12-12 2016-06-16 The Boeing Company Congruent non-uniform antenna arrays
CN104749552A (zh) * 2015-03-21 2015-07-01 西安电子科技大学 基于稀疏重构的互质阵列波达方向角估计方法
CN106896340A (zh) * 2017-01-20 2017-06-27 浙江大学 一种基于压缩感知的互质阵列高精度波达方向估计方法
CN107037392A (zh) * 2017-03-01 2017-08-11 浙江大学 一种基于压缩感知的自由度增加型互质阵列波达方向估计方法
CN107102291A (zh) * 2017-05-03 2017-08-29 浙江大学 基于虚拟阵列内插的无网格化互质阵列波达方向估计方法
CN107315160A (zh) * 2017-05-03 2017-11-03 浙江大学 基于内插虚拟阵列信号原子范数最小化的互质阵列波达方向估计方法
CN107329108A (zh) * 2017-05-03 2017-11-07 浙江大学 基于内插虚拟阵列协方差矩阵Toeplitz化重建的互质阵列波达方向估计方法
CN107589399A (zh) * 2017-08-24 2018-01-16 浙江大学 基于多采样虚拟信号奇异值分解的互质阵列波达方向估计方法
CN107422295A (zh) * 2017-08-30 2017-12-01 浙江大学 基于互质阵列虚拟域等价信号原子范数表示的波达方向估计方法
CN108344967A (zh) * 2018-01-20 2018-07-31 中国人民解放军战略支援部队信息工程大学 基于互质面阵的二维波达方向快速估计方法
CN108710102A (zh) * 2018-05-15 2018-10-26 浙江大学 基于互质阵列二阶等价虚拟信号离散傅里叶逆变换的波达方向估计方法
CN109143228A (zh) * 2018-08-28 2019-01-04 海南大学 互耦条件下双基地mimo雷达非圆目标的角度估计方法
CN109143152A (zh) * 2018-09-25 2019-01-04 哈尔滨工业大学 基于张量建模的极化阵列波达方向和极化参数估计方法
CN110244259A (zh) * 2019-06-26 2019-09-17 南京航空航天大学 数据缺失情况下基于低n-秩最小化张量填充的二维到达角估计方法
CN110927661A (zh) * 2019-11-22 2020-03-27 重庆邮电大学 基于music算法的单基地展开互质阵列mimo雷达doa估计方法
CN111610486A (zh) * 2020-05-03 2020-09-01 浙江大学 基于平面互质阵列虚拟域张量空间谱搜索的高分辨精确二维波达方向估计方法
CN111624545A (zh) * 2020-05-03 2020-09-04 浙江大学 基于结构化虚拟域张量信号处理的互质面阵二维波达方向估计方法
CN112904272A (zh) * 2021-01-18 2021-06-04 浙江大学 基于互相关张量的三维互质立方阵列波达方向估计方法
CN113552532A (zh) * 2021-07-09 2021-10-26 浙江大学 基于耦合张量分解的l型互质阵列波达方向估计方法

Also Published As

Publication number Publication date
US20240210510A1 (en) 2024-06-27
CN114444298A (zh) 2022-05-06

Similar Documents

Publication Publication Date Title
CN109655799B (zh) 基于iaa的协方差矩阵向量化的非均匀稀疏阵列测向方法
CN111610486B (zh) 基于平面互质阵列虚拟域张量空间谱搜索的高分辨精确二维波达方向估计方法
WO2021068496A1 (zh) 基于结构化虚拟域张量信号处理的互质面阵二维波达方向估计方法
WO2023279411A1 (zh) 基于耦合张量分解的l型互质阵列波达方向估计方法
JP7327840B2 (ja) 相互相関テンソルに基づく三次元の互いに素のキュービックアレイの到来方向推定方法
CN107544051A (zh) 嵌套阵列基于k‑r子空间的波达方向估计方法
WO2023137813A1 (zh) 基于最优结构化虚拟域张量填充的超分辨互质面阵空间谱估计方法
Zheng et al. Coarray tensor completion for DOA estimation
CN109613473A (zh) 基于稀疏性的展开互质线阵角度估计方法
US11422177B2 (en) Spatial spectrum estimation method with enhanced degree-of-freedom based on block sampling tensor construction for coprime planar array
CN112733327A (zh) 一种面向非高斯信号的连续和阵稀疏阵列及其设计方法
Liu et al. Direction‐of‐Arrival Estimation for Coprime Array Using Compressive Sensing Based Array Interpolation
WO2023070499A1 (zh) 基于四阶采样协方差张量去噪的子阵分置式l型互质阵列波达方向估计方法
CN109143151A (zh) 部分阵元损坏的均匀面阵张量重构方法及信源定位方法
CN106980105B (zh) 电磁矢量传感器阵列空间旋转解相干测向方法
Zheng et al. DOA estimation via coarray tensor completion with missing slices
CN109738852B (zh) 基于低秩矩阵重建的分布式源二维空间谱估计方法
US20210364591A1 (en) High-resolution, accurate, two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with co-prime planar array
WO2023137812A1 (zh) 基于虚拟域张量填充的互质面阵二维波达方向估计方法
CN110095750A (zh) 基于准平稳信号稀疏重构的快速二维欠定测角方法
CN112327292A (zh) 一种二维稀疏阵列doa估计方法
CN113687297B (zh) 数据缺失下基于矩阵分解的声矢量传感器doa估计方法
US20230048116A1 (en) Composite tensor beamforming method for electromagnetic vector coprime planar array
CN114648041A (zh) 一种基于平行稀疏阵列的二维欠定doa估计算法
Jamshidpour et al. A coarray processing technique for nested vector-sensor arrays with improved resolution capabilities

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 17920401

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22921248

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE