CN112731303B - Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application - Google Patents

Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application Download PDF

Info

Publication number
CN112731303B
CN112731303B CN202011516399.0A CN202011516399A CN112731303B CN 112731303 B CN112731303 B CN 112731303B CN 202011516399 A CN202011516399 A CN 202011516399A CN 112731303 B CN112731303 B CN 112731303B
Authority
CN
China
Prior art keywords
array
matrix
covariant
height measurement
wave radar
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN202011516399.0A
Other languages
Chinese (zh)
Other versions
CN112731303A (en
Inventor
陈根华
曾春花
莫振威
纪金程
罗晓萱
华有为
刘翠祯
江群烁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanchang Institute of Technology
Original Assignee
Nanchang Institute of Technology
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 Nanchang Institute of Technology filed Critical Nanchang Institute of Technology
Priority to CN202011516399.0A priority Critical patent/CN112731303B/en
Publication of CN112731303A publication Critical patent/CN112731303A/en
Application granted granted Critical
Publication of CN112731303B publication Critical patent/CN112731303B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO 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
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
    • G01S7/28—Details of pulse systems
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO 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
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/06—Systems determining position data of a target
    • G01S13/42—Simultaneous measurement of distance and other co-ordinates
    • G01S13/44—Monopulse radar, i.e. simultaneous lobing
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO 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
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
    • G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
    • G01S7/411—Identification of targets based on measurements of radar reflectivity
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO 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
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
    • G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
    • G01S7/414—Discriminating targets with respect to background clutter

Abstract

The invention discloses an interference array meter wave radar under non-Gaussian noise, a robust height measurement method and application thereof, wherein the height measurement method is based on a covariant matrix of fractional low-order moment of an interference array, and utilizes the covariant matrix to inhibit non-Gaussian distribution scattering components in low-angle target complex multipath reflected signals. Because the covariant matrix reserves the manifold structure of the interference array, the decorrelation of the target direct wave and the specular reflection wave can be realized by the covariant matrix and the space smoothing technology, and then the steady low-angle target height measurement can be realized by the double-scale unitary ESPRIT algorithm. According to the invention, the pitch aperture of the radar is expanded through the interference array structure, the degree of freedom of a base line is increased, and meanwhile, the non-Gaussian scattering component in the complex multipath signal is restrained through the fractional low-order moment, so that the low-angle measurement high performance of the meter wave radar is improved.

Description

Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application
Technical Field
The invention relates to the technical field of meter wave radars.
Background
The meter wave radar is an important component part of anti-stealth and anti-radiation missile and other anti-air systems, but the meter wave radar has wide beam and low positioning accuracy, and how to improve the target resolution and positioning accuracy is always an important research subject of the meter wave radar, especially the low elevation target height measurement.
The prior researches show that when the target is in a low elevation area, the wide beam irradiates the target to land, so that a strong thermal clutter phenomenon is generated, or the target generates a complex multipath signal through the reflection of the ground (sea) surface, so that the difference beam width of a single pulse is seriously widened, and the single pulse height measurement generates a larger error and even fails. Based on the above situation, the prior art mainly researches on beam width and complex multipath signals which are key factors for limiting the low elevation angle height measurement performance of the meter wave radar, such as a synthetic steering vector method, a spatial filtering method and the like for processing the beam width of the meter wave radar through a super-resolution method, and a random disturbance method of a complex multipath signal model. However, these parametric altimetric methods rely heavily on signal models, and the accuracy of the signal models directly determines the altimetric performance of the algorithm.
From Barton's multipath signal model, it is known that complex multipath signals can be separated into independent specular reflection signals (Specular Component) and scattering components (Diffuse Component), where Non-Gaussian (Non-Gaussian) distributed scattering components are difficult to model accurately, often reduced to Gaussian white noise in engineering applications, thus reducing the robustness of low angle altimetry of a meter wave radar.
On the other hand, the resolution of the array can be improved by increasing the freedom degree of the array, but the process is complex and the calculation amount is high, and the freedom degree of height measurement can be greatly reduced by a line-row synthesis technology for converting the area array into the linear array in order to reduce the calculation amount, so that the height measurement precision and resolution of the meter wave radar are limited.
Disclosure of Invention
The invention provides a meter wave radar interference array structure which can realize the expansion of array aperture, increase the freedom degree of the array and obtain high-precision low-elevation angle target height.
The invention also provides a method for stably measuring the height of the interference array meter wave radar.
The invention provides an application of an interference array meter wave radar and/or a height measurement method.
The invention firstly provides the following technical scheme:
a robust height measurement method of interference array meter wave radar in non-Gaussian noise environment comprises the following steps:
s1, snapshot data of an interference array are obtained;
s2, calculating a covariant matrix and/or covariant coefficient matrix of the fractional lower-order moment of the interference array;
s3, performing decorrelation and real valued processing on the covariant matrix and/or covariant coefficient matrix;
s4, carrying out eigenvalue decomposition on the real value covariant matrix and/or covariant coefficient matrix to obtain a signal subspace;
s5, estimating a rough estimation and a fine estimation of the cosine of the target direction by a double-scale unitary ESPRIT algorithm and automatically pairing;
s6, performing correlation method de-blurring on the paired fine estimation;
s7, obtaining the target incidence angle and/or the target height by precisely estimating the direction cosine of the deblurring.
In the above scheme, the fractional low-order moment refers to a covariant matrix with a coefficient of fractional order within 0-2.
According to some embodiments of the invention, the covariate matrixObtained by the formula:
wherein t=1, 2, …, N represents snapshot sampling time, N represents snapshot number,indicating the instant t array element is fastThe beat vector, the complex conjugate, p the order, when p=2, the fractional lower order moment is the common covariance, k, m the array element sequence number.
According to some embodiments of the invention, the covariate coefficient matrixObtained by the formula:
according to some embodiments of the invention, the decorrelation process is implemented by a two-dimensional spatial smoothing algorithm.
According to some embodiments of the invention, the real valued processing is implemented by unitary transformation.
According to some embodiments of the invention, the pairing of the coarse estimation and the fine estimation of the directional cosine is achieved by an automatic pairing algorithm.
According to some embodiments of the invention, the disambiguation is performed by an association algorithm of nearest neighbor criteria.
The invention further provides a meter wave radar interference array structure which can be applied to the height measurement method.
The invention further provides application of the height measurement method and/or the interference array meter wave radar, which is low elevation target measurement.
The measurements may include angle of incidence measurements and height measurements.
The low elevation angle is about 2 degrees or less, and is mainly aimed at a long-distance ultra-low altitude target.
The invention has the following beneficial effects:
the specific inverse T-shaped interference array provided by the invention can effectively expand the pitch array aperture of the meter wave radar, and simultaneously increase the degree of freedom of the array, thereby obtaining the target identification with high resolution and the measurement capability with high precision.
The height measurement method can restrain non-Gaussian distribution scattering components in complex multipath signals through the covariant matrix and/or covariant coefficient matrix of the fractional low-order moment of the received interference array signals.
The height measurement method can better solve the height measurement error caused by the beam width or the complex multipath signals, can expand the pitching aperture of the radar through the interference array structure, increases the baseline degree of freedom, and can inhibit the non-Gaussian scattering component in the complex multipath signals through the fractional low-order moment; the low elevation angle height measurement performance of the meter wave radar is improved.
The height measurement method simultaneously suggests that the fractional covariant matrix for inhibiting the non-Gaussian distribution scattering component of the complex multipath signal maintains the manifold structure and rank deficiency of the array as the covariance matrix, and provides a new thought for array signal processing in a non-Gaussian noise environment.
Compared with URA (Uniform Rectangular Array) with the same hardware scale, which cannot effectively distinguish the low elevation target, the interference array of the height measurement method realizes reliable distinguishing and height measurement of the low elevation target in a complex environment, and the height measurement performance of the meter wave radar can be further adjusted and optimized through optimization of fractional order coefficients of a covariant matrix.
The side elevation method can reliably distinguish and measure the target with the elevation angle below 2 degrees.
Drawings
FIG. 1 is a schematic diagram of an interferometric array meter wave radar model according to an embodiment.
Fig. 2 is a schematic diagram of a height measurement process according to an embodiment.
FIG. 3 is a graph showing the eigenvalue distribution of the fractional covariant matrix according to embodiment 1.
FIG. 4 is a graph showing the accuracy of the interferometric array estimation at different fractional lower order moments described in example 1.
Fig. 5 is a schematic diagram of the estimated direct wave and multipath signals according to embodiment 1.
Fig. 6 shows the resolution probability of the direct wave and the multipath signal according to embodiment 1.
Fig. 7 is a schematic diagram showing the effect of the spatial smoothing direction on the estimation performance in embodiment 1.
Fig. 8 is a graph showing the effect of fractional order coefficient p on altimetric performance at different SNRs as described in example 1.
FIG. 9 is a graph of e and p as described in example 1.
Detailed Description
The present invention will be described in detail with reference to the following examples and drawings, but it should be understood that the examples and drawings are only for illustrative purposes and are not intended to limit the scope of the present invention in any way. All reasonable variations and combinations that are included within the scope of the inventive concept fall within the scope of the present invention.
The target is resolved and measured by the inverted-T interference array meter wave radar shown in FIG. 1.
The interference array consists of subarrays S1 and S2, wherein S1 and S2 are uniform area arrays (URA), S1 is a main receiving and transmitting array, S2 is a receiving array only, the aperture of an array S3 in S1 is consistent with the aperture of an array S2, and the azimuth aperture of S1 is larger than S2. The array S2 is a uniform area array of Nz multiplied by Ny, the array element spacing is dy less than or equal to lambda/2, dz less than or equal to lambda/2, and the array S1 is Nz multiplied by Ny 1 Is a uniform planar array of (c). Subarrays S1 and S2 form an interference structure, the baseline of the interference array is D, and D > Nzdz.
In azimuth, the multipath signal does not broaden the azimuth difference beam width of the single pulse, so that the subarray S1 in the interference array can realize accurate tracking of the single pulse.
Based on the above-mentioned interference array, the present invention performs the target height measurement by the flow as shown in fig. 2, which includes:
s1, snapshot data of the interference array is obtained.
It may further comprise:
a three-dimensional rectangular coordinate system related to the interference array is established by taking the phase center of the array S1 as an origin, taking the array elevation as a z axis, taking a positive array surface as an x axis and taking a parallel array surface as a y axis.
Under the three-dimensional rectangular coordinate system, obtaining a snapshot vector of the interference array in the distance unit of the target to be detected through the method (1)The following are provided:
wherein u is T ＝cosθ T sin phi represents the cosine and v of the y-axis direction of the target T ＝sinθ T Representing the cosine, theta of the z-axis direction of the object T Representing the glancing angle of the target, phi representing the azimuth angle of the target; u (u) s ＝cosθ s sin phi denotes the cosine of the y-axis direction of the specular signal, v s ＝sinθ s Cosine, θ, representing the z-axis direction of the specular signal s An incident angle of a specular reflection signal representing the target, and θ s ＜0；ρ＝ρ 0 e jκΔR Representing the combined reflection coefficient of the bottom surface; ΔR is approximately equal to 2hH/R and represents the wave path difference between the direct wave and the specular reflection wave generated by the target; ρ 0 Representing the surface or sea surface reflection coefficient; s (t) represents a target signal; eta (theta) d ) Representing a scattering component energy density distribution function; θ d Representing the scattering angle; ρ d (θ d ) Representing the scattering coefficient; τ (θ) d Phi) represents the delay between the scattered component and the direct wave, n G (t) represents zero mean variance sigma 2 Complex gaussian white noise of (a); n is n d (t) represents a non-gaussian distributed scattering component; n (t) represents mixed noise (Contaminated Gaussian Model, CGM); t=1, 2, …, N denotes snapshot sampling time, N denotes snapshot number, x= [ X (1) … X (N)]。
Through this snapshot vector, under the three-dimensional rectangular coordinate system, further there is:
the interferometric array steering vectors are:
the phase center offset vector of the interference array is:
sub-area array S2 is atSteering vector of y-axis directional linear array:
steering vector of the sub-area array S2 in the z-axis direction linear array:
wherein κ=2π/λ represents wave number; lambda represents the radar wavelength; the superscript T denotes a matrix transpose,represents Kroneckner product.
On the basis of the above, the low-angle elevation method of the meter wave radar in the prior art usually ignores or simplifies the multipath scattering component into Gaussian noise so as to enable the multipath scattering component to meet the application condition of subspace algorithm, and when the reflecting surface is a complicated (sea) surface, the scattering component in the multipath signal has obvious non-Gaussian property, and the amplitude of the scattering component is related to the radar working frequency, glancing angle, shielding effect and other factors as known by the fraunenhofer criterion. Model mismatch is therefore one of the key factors in the non-robust altimetry produced by existing methods.
Unlike the prior art, the present invention further performs target height measurement by:
s2, calculating a covariate matrix and/or covariate coefficient moment of the fractional lower-order moment of the interference array.
Further, the covariate matrix C and covariate coefficient matrix Coe of the interference array are obtained by numerical averaging as follows:
wherein p is the order, x k (t) and x m (t) the k and m array element reception at time t respectivelyA signal.
Because Covariance (Covariance) of the non-Gaussian random variable is infinite, but fractional low-order moment, namely Covariance (Covariance) is bounded, and both the fractional order Covariance matrix and the Covariance matrix of the non-Gaussian random variable generated by the interference array keep manifold structures and rank deficiency of the array, a conventional subspace algorithm can be directly applied to the Covariance matrix to realize low-angle elevation, and can further realize solution correlation processing through a space smoothing method, a front-back averaging method, a Toeplitz method and the like, and then realize low-angle target angle measurement or elevation through a subspace spectrum estimation method.
Further proved as follows:
let the complex random variables X and Y of non-Gaussian distribution, its fractional low-order moment is:
C(X，Y)＝[X，Y] p ＝E[X|Y| p-2 Y * ]，0＜p＜2 (6)
where E x represents probability expectation, x represents complex conjugate, p represents fractional lower-order moment coefficient, and when p=2, fractional lower-order moment is common covariance.
Let the non-gaussian distributed complex random variables X and Y, their covariates coefficients (Covariation Coefficient) be:
based on the above settings, the fractional order covariant matrix (Covariation Matrix, CM) of the snapshot vector of the interference array is obtained asWherein element C of kth row and mth column in covariant matrix km The method comprises the following steps:
wherein x is k (t) and x m (t) is the k-th and m-th array element receiving signals at the t moment respectively.
Substituting the interference array element signals into the above formula to obtain:
wherein s is k (t)＝(a k (θ T ，φ)+ρa k (θ s ，φ))s(t)，s m (t)＝(a m (θ T ，φ)+ρa m (θ s ，φ))s(t)。
The conditional desired characteristics of probability theory and the phase rotation invariant theorem (Phase Rotational Invariance) are available:
wherein, lambda s ＝E[s(t)|(1+ρ)s(t)+n m (t)| p-2 ((1+ρ)s(t)+n m (t)) * ]Then it is a generalized signal covariate constant.
Since the mixed noise of different array elements is independent of each other, in the formula (5)The method comprises the following steps:
wherein delta km Represents the Kronecker function, γ=e [ n ] m (t)|s m (t)+n m (t)| p-2 (s m (t)+n m (t)) * ]Representing the generalized noise covariate constant.
The fractional order covariant matrix C of the interference array under the complex multipath signal is expressed as a matrix form, namely:
wherein,representing a generalized multipath signal covariant matrix,
A(θ T ，θ s ，φ)＝[a(θ T ，φ) a(θ s ，φ)]，I n representing an n-order unit array.
Obviously, the coherence of the specular signal with the direct wave is such that Λ is rank deficient.
From equation (8), CM is similar to the array covariance matrix in gaussian white noise background, namely:
R＝E[XX H ]＝A(θ T ，θ s ，φ)∑A H (θ T ，θ s ，φ)+σ 2 I 2NzNy (13)
wherein Sigma represents the signal covariance matrix, sigma 2 Representing gaussian white noise power.
From the covariate coefficient definition, the covariate coefficient matrix (Covariantion Coefficient Matrix, CCM) versus CM can be expressed as:
Coe＝C·diag[1/E(|x 1 (t)| p )，1/E(|x 2 (t)| p )，…，1/E(|x 2NzNy (t)| p )](14) Wherein diag (·) represents the diagonal matrix.
From equations (12) - (14), CM and CCM not only retain array Manifold (Manifold) information, but also effectively process non-gaussian distributed scatter component outlier points.
And S3, performing decorrelation and real valued processing on the covariate matrix and/or the covariate coefficient matrix.
It may further comprise:
a spatially smooth selection matrix of the interference array is first set.
For example, the sub-area array S2 is set to have a selected smooth sub-array number Ny along the y-axis s The z-axis aperture is unchanged, and the sub-area array S2 has L Ny s ' Nz smooth subarrays, l=ny-Ny s +1。
Similarly, the sub-area array S3 may obtain L smooth sub-arrays.
The spatial smoothing selection matrix can thus be:
wherein, the selection matrix of the first subarray in the y-axis direction
Performing two-dimensional spatial smoothing on the covariant matrix through the spatial smoothing selection matrix to obtain a smoothed covariant matrix C as follows SS ：
Wherein,representing the 2NzNy order front-back interchange matrix.
Since the interference array satisfies central symmetry (centro-symmetry), C can be further transformed (Unitary Transform) by unitary transformation SS And (3) realizing real-time value, and reducing the calculated amount to improve the real-time property of measurement.
Obtaining a smoothed real-valued covariant matrix by unitary transformation, as follows
Wherein,represents a sparse Unitary Matrix (Unitary Matrix), and
the Forward-backward average (Forward/backward Averaging) in the real valued process can further improve the decorrelation performance of the covariant matrix, and create conditions for realizing stable height measurement by the unitary ESPRIT algorithm.
S4, carrying out eigenvalue decomposition on the real value covariant matrix and/or covariant coefficient matrix to obtain a signal subspace;
the real value covariant matrix is subjected to eigenvalue decomposition to obtain
Wherein E is s And E is n Representing signal and noise subspaces, E s Is composed of two feature vectors with maximum feature values, E n Then it is composed of the remaining feature vectors,diagonal matrix of two largest eigenvalues ++>Representing a diagonal matrix of residual eigenvalues.
S5, estimating a rough estimation and a fine estimation of the cosine of the target direction by an ESPRIT algorithm and automatically pairing.
Specifically, two nzxNy are obtained after spatial smoothing s The interference array formed by subarrays has displacement invariance of offset dz in two subarrays on the z axis, and has displacement invariance of offset D between subarrays. The double-scale ESPRIT algorithm can obtain low-precision but non-fuzzy directional cosine estimation, called coarse estimation, by the motion invariance of the offset dz; whereas a highly accurate but cycle-blurred direction cosine estimate is obtainable from the shift invariance of the offset D, called the fine estimate.
Further, by unitary ESPRIT algorithm, the shift invariance for offset dz and D can be expressed as:
K c1 [a r (u T ，υ T )a r (u s ，υ s )]Φ c ＝K c2 [a r (u T ，υ T )a r (u s ，υ s )] (20)
K f1 [a r (u T ，υ T )a r (u s ，υ s )]Φ f ＝K f2 [a r (u T ，υ T )a r (u s ，υ s )] (21)
wherein,and->Selection matrix phi representing coarse and fine estimates, respectively c And phi is f Representing real-valued rotation operator +_>Representing subarray real value guide vector,/>Guiding vector representing the spatially smoothed interferometric matrix, < >>Is a y-axis arrayIs a coarse estimate and a fine estimate of the target spatial frequency of +.>And->Coarse and fine estimates of the spatial frequency of the specular reflection signal are +.>And->Re[·](Im[·]) Representing the real (imaginary) part.
From the periodicity of the tangent function it is known that,and->Blur may occur.
From the ESPRIT algorithm, the presence of the non-singular matrix T, the rotational invariance for offsets dz and D, can be expressed as:
wherein the operator ψ is rotated c ＝TΦ c T -1 And psi is f ＝TΦ f T -1 Can be solved by the total least squares method (TLS).
To psi c And psi is f After the eigenvalue decomposition, the direction cosine coarse estimation and the direction cosine fine estimation can be obtained, but the one-to-one correspondence relation cannot be ensured, so that correct pairing is needed before the fine estimation is deblurred.
Specifically, the rotation operator ψ obtained by the real valued processing c And psi is f Is real-valued and can be obtained by an automatic pairing algorithm, e.g. document PILLAI S U and KWON B H.forward/backward spatial smoothing techniques for coherent signal identification [ J]IEEE Trans. On operational, speech and Signal Processing, constructing complex-valued rotation operator ψ to achieve accurate pairing, i.e
Ψ＝Ψ c +jΨ f ＝T(Φ c +jΦ f )T -1 (28)
Wherein ψ is c And psi is f Representing the real and imaginary parts, respectively.
The rough estimation and the fine estimation of the paired direction cosine are respectively as follows:
wherein, gamma k The eigenvalues of the operator ψ are rotated for complex values.
And S6, performing correlation method defuzzification on the paired fine estimation.
Specifically, since the base line D > Nzdz > λ, the periodicity of the tangent function, i.eIt is known that the direction cosine fine estimate may be ambiguous, i.e.:
where k is the wave number and n is an integer.
From the correlation method of nearest neighbor criteria, the precise estimation is deblurred by taking the rough estimation without ambiguity as a reference, namely
Wherein, representing an upward (downward) rounding.
S7, obtaining the target incidence angle and/or the target height by precisely estimating the direction cosine of the deblurring.
Specifically, after the direction cosine fine estimation deblurring, the high-precision non-blurring incident angle of the target can be obtained by the following formula,
by passing throughThe target height is obtained, where R is the target distance.
The prior art section referred to in the following examples refers to the following technical literature:
XU W C,CHEN C R,and DAI J S.Detection of known signals in additive impulsive noise based on Spearman’s rho and Kendall’s tau[J].Signal processing,2019.
ZUO Weiliang,XIN Jingmin,and LIU Wenyi.Localization of Near-Field Sources Based on Linear Prediction and Oblique Projection Operator[J].IEEE Trans.On Signal Processing,2019.
example 1
And carrying out simulation experiments on the interference array and the height measurement method in the specific embodiment.
Wherein nz=8, ny=10, n=20, h=30m, h=3000 m, r=100 km, ρ is set 0 ＝-0.97，d＝0.5λ，λ＝1.2m，θ T ，θ s ]＝[1.703°，-1.74°]100 Monte Carlo tests were performed on each data point, SNR is defined as the signal-to-noise ratio of the array element, and a common epsilon-mixed noise model (CGM) is selected to simulate non-Gaussian distribution of scattered components and white noise, and the probability density function is as described in the reference document
f＝(1-ε)N c (0，σ 2 I 2NzNy )+εN c (0，(kσ) 2 I 2NzNy ) Wherein N is c (0，σ 2 I 2NzNy ) Represents zero-mean normal distribution probability density distribution function (PDF), epsilon (0, 1) is binomial distribution probability, and epsilon mixed noise power is P n ＝(1-ε)σ 2 +εk 2 +σ 2 ，k＝30。
On the basis of the above, experiments 1 to 7 were performed, in which "CM" represents a covariant matrix, "CCM" represents a covariant coefficient matrix, "SCM" represents a covariance matrix, "SS" represents spatial smoothing, and "URA" represents a uniform area matrix.
The following are provided:
test 1
And verifying the rank deficiency of the interference arrays CM and CCM and the effectiveness of a two-dimensional space smoothing decoherence algorithm when the target is at a low elevation angle.
Under the above simulation conditions, snr=0db, d=15λ was set, and smoothed three times along the azimuth dimension. CM, CCM and SCM of 100 times of interference array as shown in fig. 3 and the first three characteristic value distribution situations after spatial smoothing can be obtained.
It can be seen that the multipath reflection signals make CM and CCM of the interference array show rank deficiency, and the two-dimensional space smoothing method can effectively realize direct and multipath signal decoherence, and recover Subspace structures of the two, wherein the CCM smoothing decoherence effect when p=0.6 is particularly remarkable, and the Subspace interchange (Subspace Swap) probability based on Subspace height measurement algorithm is greatly reduced. Meanwhile, the fractional order covariant theory is suitable for processing scattering components with non-Gaussian distribution, and the robustness of the height measurement algorithm is guaranteed. The non-gaussian distributed scattering components do not preserve the array manifold information and therefore the SCM and spatial smoothing algorithms are not suitable for non-gaussian distributed multipath signal environments.
Test 2
And verifying the validity of the interference array and the height measurement method.
Under the above simulation conditions, the Root Mean Square Error (RMSE) of the low-angle target angle measurement based on the fractional covariate matrix at different baselines is shown in fig. 4, wherein
As can be seen from fig. 4, at low SNR, the measurement performance based on the fractional lower moment is significantly better than that of the conventional SCM, and at p=0.6, the measurement accuracy based on CCM is about 5 times that of the SCM, which is far better than that of the conventional SCM method, and it is fully explained that the fractional lower moment is effective in processing the scattered component of the non-gaussian distribution.
Meanwhile, fig. 4 also shows that in the high resolution region, the angular accuracy is improved by about 4 times when d=201 than when d=81, wherein the estimated performance of CCM is best when p=0.6, and CM times when p=1.2.
The height measurement performance of the d=81 and 201 interferometric arrays improves by up to 4 and 6 times, respectively, compared to a URA of 10'16 of equivalent hardware scale, and greatly reduces the SNR threshold of the algorithm.
The test fully demonstrates the robustness of the fractional low-order moment to the non-Gaussian distribution scattering component and the high-resolution robust height measurement of the interference array can be realized.
Test 3
And analyzing the high resolution performance of the interference array and the height measurement method.
For closed estimation algorithms such as ESPRIT and root-finding MUSIC, the target is often considered to be resolved when the estimation error is smaller than the half-wave beam width, i.eAnd->
Under the above simulation conditions, fig. 5 shows the direct and specular signal distribution of the interference array and URA estimation at snr=0 dB. Obviously, three kinds of height measurement methods based on URA can not realize reliable resolution of direct wave and multipath signals, and the interference array can realize reliable resolution. From fig. 6, it can be seen that the resolution performance based on CCM also tends to be 100% at low SNR, while the resolution performance of CM is much lower than CCM, but the resolution performance of the height measurement based on fractional low-order moment is significantly better than that of the conventional SCM method. The interference array and the height measurement method make full use of the degree of freedom of the interference array, and improve the low-angle target resolution of the meter wave radar.
Test 4
And verifying the influence of the space smoothing direction and the times on the performance of the height measurement method.
Fig. 7 shows the height measurement accuracy at different smoothing directions and times at d=10λ. The conventional SCM and CM are favored for their altimetric performance when smoothed along the azimuth dimension, keeping the array pitch aperture unchanged, while CCM performs better when smoothed along the pitch dimension. As can be seen from fig. 7, the height measurement performance of the CCM after pitch smoothing is improved by about 3 times compared with the azimuth smoothing direction, so that the height measurement performance can be optimized when the CCM needs to be smoothed and decorrelated, but the CCM and SCM need to be smoothed in the azimuth direction.
Test 5
And analyzing the performance influence of the fractional order coefficient p on the interference array and the height measurement method.
As can be seen from fig. 8, the CM has a p-threshold window with p=0.8 as the center and a width of 0.2, and has stable estimation performance outside the p-threshold window, and is insensitive to SNR; the CCM has obvious p threshold, the low-angle target can be reliably distinguished when the CCM is lower than the threshold, and the p threshold is increased along with the increase of SNR, namely, the p value is required to be optimally selected according to the characteristic of non-Gaussian distribution scattering components when the array covariant matrix is calculated, so that the performance of a subsequent height measurement algorithm is optimal. The present invention selects p=1.2 and p=0.6 when calculating CM and CCM, respectively.
Test 6
Analysis of the relation of the mixed noise at different epsilon and the fractional order coefficient p.
FIG. 9 is a graph of RMSE contours obtained from CM and CCM at different e, where SNR=0 dB, ny s =8. As can be seen from FIG. 9, when CM is used, different ε have corresponding p-threshold windows, and the width of the threshold windowsThe degree increases with increasing e; but when p>1, is suitable for arbitrary epsilon-mixed noise, therefore, engineering applications often choose p=1.2 to obtain robust altimetric performance. However, the CCM-based altimetric performance has an approximately symmetric distribution relationship with respect to ε and p, i.e., when ε is large, p is selected to be around 2, and when ε < 0.6, p is selected<1. The performance is better when p=0.6 is selected in combination with the non-gaussian distribution characteristic of multipath scattering components of the low-angle target of the meter wave radar. In addition, this test again demonstrates that CCM-based altimetric performance is superior to CM.
The above examples are only preferred embodiments of the present invention, and the scope of the present invention is not limited to the above examples. All technical schemes belonging to the concept of the invention belong to the protection scope of the invention. It should be noted that modifications and adaptations to the present invention may occur to one skilled in the art without departing from the principles of the present invention and are intended to be within the scope of the present invention.

Claims (8)

1. The robust height measurement method of the interference array meter wave radar under the non-Gaussian noise is characterized by comprising the following steps of: the method comprises the following steps:
s1, acquiring snapshot data of an interference array;
s2, calculating a fractional low-order moment covariant matrix and/or covariant coefficient matrix of the interference array;
s3, performing decorrelation and real valued processing on the covariant matrix and/or covariant coefficient matrix to obtain a real value covariant matrix;
s4, carrying out eigenvalue decomposition on the real value covariant matrix to obtain a signal subspace;
s5, estimating rough estimation and fine estimation of the cosine of the target direction by a double-scale unitary ESPRIT algorithm and automatically pairing;
s6, performing correlation method de-blurring on the paired fine estimation;
s7, obtaining an incident angle and/or a target height by precisely estimating the direction cosine of the deblurred object;
wherein the covariant matrixObtained by the formula:
the covariate coefficient matrixObtained by the formula:
where t=1, 2,..n represents the snapshot sampling time, N represents the snapshot number, x (t) represents the snapshot vector of the array element at the time t, representing complex conjugate, p represents fractional order coefficient, k and m represent array element serial numbers.
2. The robust altimetric method of claim 1, where: the decoherence processing is realized by a two-dimensional space smoothing algorithm.
3. The robust altimetric method of claim 1, where: the real valued processing is implemented by unitary transformation.
4. The robust altimetric method of claim 1, where: the pairing is achieved by an automatic pairing algorithm.
5. The robust altimetric method of claim 1, where: the solution ambiguity is achieved by a nearest neighbor criterion association algorithm.
6. The robust altimetric method of any one of claims 1 to 5, characterized by: the fractional lower order moment coefficient p=0.6-1.2.
7. An interferometric array meter wave radar employing the robust altimetric method of any one of claims 1 to 6.
8. Use of the altimetric method of any one of claims 1 to 6 or the interferometric array meter wave radar of claim 7 for identifying and/or measuring low elevation targets.
CN202011516399.0A 2020-12-21 2020-12-21 Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application Active CN112731303B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011516399.0A CN112731303B (en) 2020-12-21 2020-12-21 Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011516399.0A CN112731303B (en) 2020-12-21 2020-12-21 Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application

Publications (2)

Publication Number Publication Date
CN112731303A CN112731303A (en) 2021-04-30
CN112731303B true CN112731303B (en) 2024-04-09

Family

ID=75603646

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011516399.0A Active CN112731303B (en) 2020-12-21 2020-12-21 Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application

Country Status (1)

Country Link
CN (1) CN112731303B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104155629A (en) * 2014-08-07 2014-11-19 哈尔滨工程大学 Method for estimating signal DOA (direction of arrival) under fewer snapshots and impulsive noise background
CN108120953A (en) * 2017-12-27 2018-06-05 大连理工大学 A kind of radio location method based on Mutual coupling
CN108303683A (en) * 2018-01-29 2018-07-20 西安邮电大学 Single not rounded signal angle methods of estimation of base MIMO radar real value ESPRIT

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102008010882A1 (en) * 2008-02-25 2009-09-03 IAD Gesellschaft für Informatik, Automatisierung und Datenverarbeitung mbH Device and method for direction estimation and / or decoding of secondary radar signals
US10229092B2 (en) * 2017-08-14 2019-03-12 City University Of Hong Kong Systems and methods for robust low-rank matrix approximation

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104155629A (en) * 2014-08-07 2014-11-19 哈尔滨工程大学 Method for estimating signal DOA (direction of arrival) under fewer snapshots and impulsive noise background
CN108120953A (en) * 2017-12-27 2018-06-05 大连理工大学 A kind of radio location method based on Mutual coupling
CN108303683A (en) * 2018-01-29 2018-07-20 西安邮电大学 Single not rounded signal angle methods of estimation of base MIMO radar real value ESPRIT

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
DOA tracking based on MIMO radar in impulsive noise background;Ming Diao等;《2014 12th International Conference on Signal Processing（ICSP）》;摘要，第263页第2栏，第264页第1栏 *
基于Screened Ratio原理的冲击噪声环境下DOA估计算法;何劲等;电子与信息学报(第05期);109-112 *
干涉阵列米波雷达的低仰角高精度估计方法;陈根华等;西安电子科技大学学报(第06期);第42-48页及图1-3 *
用干涉式APES算法实现干涉阵盲DOA估计;朱伟等;西安电子科技大学学报(第01期);195-202 *

Also Published As

Publication number Publication date
CN112731303A (en) 2021-04-30

Similar Documents

Publication Publication Date Title
Melvin et al. An approach to knowledge-aided covariance estimation
Touzi et al. Coherence estimation for SAR imagery
US7667639B2 (en) Passive channel calibration method based on non-linear antenna array
CN107870314B (en) Complete electromagnetic component weighting fusion direction-finding optimization method based on polarization sensitive array
Yuan Coherent source direction-finding using a sparsely-distributed acoustic vector-sensor array
Zhao et al. Altitude measurement of low elevation target based on iterative subspace projection
Chen et al. Short-range clutter suppression for airborne radar using sparse recovery and orthogonal projection
Yang et al. Enhanced knowledge-aided space–time adaptive processing exploiting inaccurate prior knowledge of the array manifold
Liu et al. Target localization in high-coherence multipath environment based on low-rank decomposition and sparse representation
Moore et al. Three-dimensional position accuracy in circular synthetic aperture radar
CN112731303B (en) Interference array meter wave radar under non-Gaussian noise and steady height measurement method and application
CN112881971B (en) Direction finding method for coherent interference source under electromagnetic directional mutual coupling effect
CN115166725A (en) Meter-wave polarization sensitive array radar height measurement method based on MUSIC algorithm
Schmieder et al. Interference cancellation and signal direction finding with low complexity
CN113820654A (en) S-band radar target low elevation DOA estimation method based on beam domain dimension reduction
Zhang et al. DOA estimation method for underwater acoustic signals based on two-dimensional power distribution (TPD) for few element array
Vasile et al. Stable scatterers detection and tracking in heterogeneous clutter by repeat-pass SAR interferometry
Sun et al. Phase array amplitude and phase correction for HFSWR based on AIS information
Galy et al. Joint detection estimation problem of monopulse angle measurement
Ly et al. Array processing application: angular superresolution for scanning antenna
Tang et al. Bistatic MIMO radar height estimation method based on adaptive beam-space RML data fusion
Sebt et al. Geometric Arithmetic Mean Method for Low Altitude Target Elevation Angle Tracking
CN113655458B (en) Space-time adaptive processing method and device based on dictionary correction and storage medium
Berry et al. Real aperture imaging using sparse optimisation with application to low-angle tracking
Bialer et al. A Multi-radar Joint Beamforming Method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant