CN-121978639-A - Non-uniform linear array MIMO radar parameter estimation method based on phase compensation

CN121978639ACN 121978639 ACN121978639 ACN 121978639ACN-121978639-A

Abstract

The invention provides a non-uniform linear array MIMO radar parameter estimation method based on phase compensation, and relates to the technical field of array signal processing. The method comprises the steps of S1, modeling a received signal, calculating a covariance matrix of the received signal, carrying out eigenvalue decomposition, S2, constructing a virtual matrix selection matrix, establishing a space translation relation, solving a twiddle factor matrix of a corresponding matrix, S3, extracting a phase principal value of eigenvalue of the twiddle factor matrix to obtain an observed phase difference vector containing phase ambiguity, S4, selecting a non-ambiguous adjacent array element pair set, predicting a real phase difference, carrying out phase compensation on the observed phase difference, and calculating to obtain a DOA and DOD accurate estimated value of a target, S5, further estimating a polarization parameter, thereby completing joint estimation of the target polarization parameter, solving the technical problems of phase ambiguity of a non-uniform array, coupling of polarization and space dimension, improving the accuracy of angle estimation, and simultaneously taking into account the accurate estimation of the target polarization parameter.

Inventors

WEN FANGQING
WANG QINWEN
ZHANG HONG

Assignees

三峡大学

Dates

Publication Date: 20260505
Application Date: 20260122

Claims (10)

1. The non-uniform linear array MIMO radar parameter estimation method based on phase compensation is characterized by comprising the following steps of: S1, modeling a received signal of a bistatic MIMO radar system formed by M unevenly distributed transmitting sensors and N unevenly distributed receiving sensors, calculating a covariance matrix of the received signal, and carrying out eigenvalue decomposition on the covariance matrix to obtain a signal subspace and a noise subspace; S2, respectively constructing a selection matrix of a transmitting dimension and a receiving dimension according to the array element position distribution of the transmitting array and the receiving array, extracting a corresponding subspace block from the signal subspace by utilizing the selection matrix, establishing a space translation relation based on the subspace block, and solving to obtain a twiddle factor matrix of the transmitting dimension and the receiving dimension; s3, extracting a phase main value of the twiddle factor matrix eigenvalue to obtain an observation phase difference vector containing phase ambiguity; S4, selecting a non-fuzzy adjacent array element pair set based on array element spacing information, predicting a real phase difference by utilizing a linear relation between a non-fuzzy phase observation value corresponding to the set and an array element spacing difference, determining the whole period number of phase winding by comparing the predicted phase difference with an actual measured phase difference, performing phase compensation on the observed phase difference, and calculating a DOA and DOD accurate estimation value of a target by utilizing the compensated phase; S5, respectively extracting subspace components corresponding to the two orthogonal polarization channels from the subspace of the virtual array signal based on the DOA and DOD accurate estimation values, constructing a polarization rotation factor based on a least square principle, decomposing the characteristic value of the polarization rotation factor, and respectively estimating a polarization pitch angle and a polarization phase difference of the target according to the amplitude and the phase of the characteristic value.
2. The non-uniform linear array MIMO radar parameter estimation method based on phase compensation according to claim 1, wherein in step S1, D array element distances between the transmitting sensor and the receiving sensor are larger than half wavelength; Under a narrow band assumption, the sensor array spatial response vector is transmitted And receiving a sensor array spatial response vector The method comprises the following steps of: ; ; Wherein, the Indicating the location of the transmitting array element, Indicating the location of the transmitting array element, Is the wavelength of the signal; For the departure direction angle DOD of the kth target relative to the transmit array, Is the reception direction angle DOA relative to the reception array; modeling the received signals of the K targets as; ; Wherein, the , Represents the Khatri-Rao product; in order to transmit the array response matrix, In order to receive the array response matrix, As a source signal, a signal is provided, Is a received Gaussian white noise matrix; polarization vector representation Wherein: ; ; The covariance matrix of the received signal is expressed as: ; Wherein, the Is a signal covariance matrix.
3. The phase compensation-based non-uniform linear array MIMO radar parameter estimation method according to claim 1, wherein in step S1, the expression for performing feature decomposition on the covariance matrix is: ; Wherein, the Is that The diagonal elements of which contain K larger eigenvalues, A matrix formed by the eigenvectors corresponding to the K larger eigenvalues; contains 2MN-K smaller eigenvalues, A matrix formed by the eigenvectors corresponding to the residual smaller eigenvalues; As seen as a signal subspace, Considered as noise subspaces; signal subspace The same as the subspace spanned by the space response vector matrix a, namely: ; Wherein the matrix Is a nonsingular full order matrix.
4. The phase compensation-based non-uniform linear array MIMO radar parameter estimation method according to claim 1, wherein in step S2, constructing a transmission dimension and reception dimension selection matrix includes: The following selectivity matrix is defined: ; ; Wherein, the Is that Is arranged in the row m of the table (a), Is that Is arranged in the row n of the (c), Is that Is used for the matrix of units of (a), Is that Is used for the matrix of units of (a), Is that Is a matrix of units of (a); And obtaining a signal subspace block corresponding to the array element through the operation of the selective matrix and the signal subspace, and further obtaining a corresponding twiddle factor matrix through calculation.
5. The method for estimating parameters of non-uniform linear array MIMO radar based on phase compensation according to claim 1, wherein in step S3, the rotation factor matrix of the transmitting dimension and the receiving dimension is subjected to feature decomposition to obtain respective feature values, the phase principal values thereof are extracted, and all pairs of adjacent array elements are stacked to correspond to the phase principal values, thereby obtaining an observed phase difference vector containing phase ambiguity.
6. The phase compensation-based non-uniform linear array MIMO radar parameter estimation method according to claim 1, wherein in step S4, the condition of selecting the non-fuzzy adjacent pair set is that the distance between the adjacent array element pairs is less than or equal to half wavelength; constructing vectors of all adjacent pitches meeting the conditions Its corresponding observed phase constitutes a vector And (3) with The linear relationship is obtained as: 。
7. the phase compensation-based non-uniform linear array MIMO radar parameter estimation method of claim 6, wherein in step S4, the linear relationship obtains a predicted value And (3) with ; ; Determining the number of complete periods of phase wrapping by comparing the predicted phase difference with the actual measured phase difference And (3) with The expression is: ; Wherein, the To round up the rounding function.
8. The phase compensation-based non-uniform linear array MIMO radar parameter estimation method according to claim 1, wherein in step S4, compensating the observed phase comprises: Based on the number of phase winding cycles And Compensating the original measured phase to obtain a compensated phase difference And : ; By using the compensated phase, DOA and DOD of the signal can be finally calculated, and the expression is as follows: 。
9. The phase compensation-based non-uniform linear array MIMO radar parameter estimation method according to claim 1, wherein in step S5, two polarization selection matrices are constructed And : ; ; Wherein, the Is an identity matrix of mn×mn; from the signal subspace by means of a polarization selection matrix Extracting subspace components corresponding to the first polarization channel and the second polarization channel: ; ; Under the principle of least square, the polarization rotation factor is constructed as The method comprises the following steps: ; Decomposing the characteristic value to obtain a corresponding characteristic value 。
10. The non-uniform linear array MIMO radar parameter estimation method based on phase compensation according to claim 1, wherein in step S5, a polarization rotation factor is obtained and eigenvalue decomposition is performed, and a polarization pitch angle and a polarization phase of a target are estimated from an amplitude and a phase of the eigenvalue, respectively, specifically: Under the adopted polarization response model, the following relation is satisfied between the characteristic value and the target polarization parameter: ; Wherein, the Representing the polarization pitch angle of the target, Indicating the polarization phase difference.

Description

Non-uniform linear array MIMO radar parameter estimation method based on phase compensation Technical Field The invention relates to the technical field of array signal processing, in particular to a non-uniform linear array MIMO radar parameter estimation method based on phase compensation. Background Along with the development of radar detection technology, polarization information is widely focused in the fields of target identification, classification, parameter inversion and the like because the polarization information can represent the electromagnetic scattering characteristics of targets. In a dual-base multiple-input multiple-output (MIMO) radar system, a high-dimensional virtual array may be formed by the cooperative operation of a transmit array and a receive array, enabling joint estimation of a target transmit direction angle (Direction of Departure, DOD) and a receive direction angle (Direction of Arrival, DOA). Meanwhile, the introduction of polarization sensitive array elements enables estimation of target polarization parameters to be possible, and important information support is provided for refined perception. In the prior art, the parameter estimation method for the bistatic polarization MIMO radar mainly faces the following challenges that the existing parameter estimation method based on subspace is divided into two types, one type is a high-resolution algorithm based on spectrum search, spectrum peak search is needed in space, the calculation complexity is extremely high, performance degradation is easy to occur under the condition of low signal-to-noise ratio or limited snapshot number, and the other type is an ESPRIT algorithm, and the method does not need spectrum search and is high in calculation efficiency, but is limited to a uniform linear array or a monostatic model, and cannot adapt to a non-uniform array structure adopted in an actual system for reducing hardware cost and meeting layout limitation. 2. Although the search-free algorithm based on rotation invariance (such as ESPRIT) is efficient in calculation, a theoretical model of the search-free algorithm generally requires an array to have a uniform or specific geometric structure (such as a uniform linear array), the array element spacing is generally limited to be within half wavelength so as to avoid the problem of phase ambiguity, the flexibility of array design is limited in actual engineering, and the search-free algorithm is difficult to adapt to the requirements of low cost, sparse or non-uniform array. 3. The polarization dimension and the space dimension have a strong coupling relation, and the polarization parameter can be estimated by introducing an additional spectrum searching step or a complicated parameter decoupling flow in the conventional method, so that the algorithm complexity is increased, and the estimation precision and efficiency are reduced. 4. The existing partial algorithm (such as a method based on Tensor Ring Decomposition (TRD)) needs to additionally increase tensor decomposition steps when estimating polarization parameters, has poor model expansibility and limited engineering applicability, and cannot meet the actual requirements of high-dimensional parameter efficient estimation under a non-uniform array. Disclosure of Invention The invention mainly aims to provide a non-uniform linear array MIMO radar parameter estimation method based on phase compensation, which solves the technical problems of high computational complexity, strict requirements on array element arrangement, high coupling and complexity in a parameter estimation process and low angle and polarization parameter estimation efficiency and accuracy. In order to solve the technical problems, the technical scheme adopted by the invention is that the non-uniform linear array MIMO radar parameter estimation method based on phase compensation comprises the following steps: S1, modeling a received signal of a bistatic MIMO radar system formed by M unevenly distributed transmitting sensors and N unevenly distributed receiving sensors, calculating a covariance matrix of the received signal, and carrying out eigenvalue decomposition on the covariance matrix to obtain a signal subspace and a noise subspace; S2, respectively constructing a selection matrix of a transmitting dimension and a receiving dimension according to the array element position distribution of the transmitting array and the receiving array, extracting a corresponding subspace block from the signal subspace by utilizing the selection matrix, establishing a space translation relation based on the subspace block, and solving to obtain a twiddle factor matrix of the transmitting dimension and the receiving dimension; s3, extracting a phase main value of the twiddle factor matrix eigenvalue to obtain an observation phase difference vector containing phase ambiguity; S4, selecting a non-fuzzy adjacent array element pair set based on array element spacing information, pr