阅读说明：本技术 一种多径信号波达方向和扩散角快速估计新方法 (New method for quickly estimating direction of arrival and diffusion angle of multipath signal ) 是由 田全 蔡睿妍 于 2021-06-28 设计创作，主要内容包括：本发明公开了一种多径信号波达方向和扩散角快速估计新方法,具体包括：构建阵列输出信号的分布源模型,有效解决点源模型下波达方向估计精度低的问题；通过噪声子空间构建二维空间谱,在所述二维空间谱进行分布源波达方向和扩散角的估计；采用单峰共轭对称函数对分布源的确定性角信号密度函数进行描述,并结合小角度近似理论,重建均匀线性阵列的广义阵列流形,实现波达方向和扩散角的分离；构建二次优化函数,得到波达方向的一维空间谱,通过该一维空间谱获得波达方向估计；获取扩散角的一维空间谱,通过该一维空间谱获得扩散角估计。本发明提升了算法可靠性,降低了算法的计算复杂度,能够适应于对实时性要求较高的无线电监测与定位场景。(The invention discloses a new method for quickly estimating the direction of arrival and the spread angle of multipath signals, which specifically comprises the following steps: a distributed source model of array output signals is constructed, and the problem of low accuracy of estimation of the direction of arrival under a point source model is effectively solved; constructing a two-dimensional space spectrum through a noise subspace, and estimating the arrival direction and the diffusion angle of a distributed source wave in the two-dimensional space spectrum; describing a deterministic angle signal density function of a distribution source by adopting a unimodal conjugate symmetric function, and reconstructing a generalized array manifold of a uniform linear array by combining a small angle approximation theory to realize the separation of the direction of arrival and the diffusion angle; constructing a quadratic optimization function to obtain a one-dimensional space spectrum of the direction of arrival, and obtaining estimation of the direction of arrival through the one-dimensional space spectrum; and acquiring a one-dimensional spatial spectrum of the diffusion angle, and acquiring diffusion angle estimation through the one-dimensional spatial spectrum. The method improves the reliability of the algorithm, reduces the calculation complexity of the algorithm, and can be suitable for radio monitoring and positioning scenes with high real-time requirements.)

1. A new method for fast estimation of the direction of arrival and the spread angle of a multipath signal is characterized by comprising the following steps:

constructing a distributed source model of the array output signals;

acquiring a covariance matrix of array output signals, performing singular value decomposition on the covariance matrix to obtain a signal subspace and a noise subspace, constructing a two-dimensional space spectrum through the noise subspace, and estimating the direction of arrival and the diffusion angle of a distributed source in the two-dimensional space spectrum;

describing a deterministic angle signal density function of a distribution source by adopting a unimodal conjugate symmetric function, and reconstructing a generalized array manifold of a uniform linear array by combining a small angle approximation theory to realize the separation of the direction of arrival and the diffusion angle;

constructing a quadratic optimization function to obtain a one-dimensional space spectrum of the direction of arrival, and obtaining estimation of the direction of arrival through the one-dimensional space spectrum;

and acquiring a one-dimensional spatial spectrum of the diffusion angle, and acquiring diffusion angle estimation through the one-dimensional spatial spectrum.

2. The new method for fast estimation of the direction of arrival and the spread angle of a multipath signal according to claim 1, wherein L distributed source signals are incident on a uniform linear array, the uniform linear array comprises M signal receiving sensors, and a signal vector x output by the sensors is expressed as:

wherein, for the ith distributed source, s_i(θ,ψ_i) In order to be the angular signal density,direction of arrival including target sourceAnd the angle of divergence sigma_iA (θ) is a guide vector of the uniform linear array, n is a noise vector, and Θ ═ π/2, π/2]Is the observation range of the distributed source;

since the signal and noise are uncorrelated, the correlation matrix for the signal vector x is represented as:

R_xx＝E(xx^H)＝R_ss(ψ)+R_nn (2)

wherein R is_ss(psi) is the correlation matrix of the clean signal, R_nnA correlation matrix that is noise;

the angular cross-correlation kernel is defined as:

then R is_ss(ψ) is rewritten as:

since multiple target sources are typically uncorrelated, the angular cross-correlation kernel is simplified to:

ρ_ij(θ,θ′,ψ_i,ψ_j)＝δ_ij·μ_i(θ,θ′,ψ_i) (5)

wherein, delta_ijIs a Kroneckerdelta function and the angular autocorrelation kernel for the ith target source is:

at this time, the angular signal density of the distributed source is:

s_i(θ,ψ_i)＝χ(θ,ψ_i)ζ_i (7)

wherein, x (theta, psi)_i) As a function of the deterministic angular signal density, ζ_iIs a complex random variable.

3. The new method for rapidly estimating the direction of arrival and the spread angle of the multipath signal according to claim 1, wherein a covariance matrix of an array output signal is obtained, and singular value decomposition is performed on the covariance matrix to obtain a signal subspace and a noise subspace, and the method specifically comprises the following steps:

for M characteristic values fromSorting from large to small, dividing the eigenvectors corresponding to L larger eigenvalues into one group, dividing the eigenvectors corresponding to the other M-L smaller eigenvalues into another group, and then taking the eigenvectors corresponding to the L larger eigenvalues as the signal subspace U_sThe eigenvectors corresponding to the M-L smaller eigenvalues are the noise subspace U_n。

4. The new method for rapidly estimating the direction of arrival and the spread angle of the multipath signal according to claim 3, wherein a two-dimensional space spectrum is constructed by the noise subspace, specifically:

estimating the arrival direction and the diffusion angle of the distributed source in the two-dimensional space spectrum, specifically:

wherein the content of the first and second substances,b (ψ) is a generalized steering vector for a uniform linear array, which is an estimate of ψ.

5. The new method for fast estimation of direction of arrival and spread angle of multipath signal as claimed in claim 1, wherein the deterministic angular signal density function of the distributed source is described by a unimodal conjugate symmetric function:

at this time, in combination with the small angle approximation theory, the generalized array manifold for reconstructing a uniform linear array is:

the function h (σ) is defined as:

thus, the generalized array manifold for a uniform linear array is further expressed as:

equation (13) achieves separation of the direction of arrival and the divergence angle.

6. The new method for rapidly estimating the direction of arrival and the spread angle of the multipath signal according to claim 1, wherein a quadratic optimization function is constructed, specifically:

reconstructing the two-dimensional spatial spectrum of the distributed source using equation (13):

defining functionsComprises the following steps:

then

Constructing quadratic optimization functionComprises the following steps:

7. the new method for rapidly estimating the direction of arrival and the spread angle of the multipath signal according to claim 6, wherein a one-dimensional spatial spectrum of the direction of arrival is obtained, and the estimation of the direction of arrival is obtained through the one-dimensional spatial spectrum, specifically:

to ensure that the solution of the quadratic optimization function in equation (17) is not zero, the quadratic optimization function is rewritten as the following constraint form:

defining functionsComprises the following steps:

solving the partial derivative with respect to h (σ):

the solution for f (θ, σ) is then:

the one-dimensional spatial spectrum of the direction of arrival estimate is:

performing a one-dimensional spectral peak search on equation (22) to obtain a direction of arrival estimate

8. A new method for fast estimation of the direction of arrival and spread of a multipath signal according to claim 4 or 7, characterised in that the method is applied to the estimation of the direction of arrival and spread of a multipath signalAnd (9) carrying out equation (9), acquiring a one-dimensional space spectrum of the diffusion angle, and performing spectrum peak search on the one-dimensional space spectrum to obtain diffusion angle estimation.

Technical Field

The invention relates to the field of passive radio monitoring and positioning, in particular to a novel method for quickly estimating the direction of arrival and the diffusion angle of multipath signals.

Background

The beginning of array signal processing technology dates back to the adaptive antenna technology in the 40 s of the 20 th century where phase-locked loops were used for antenna tracking. In 1965, the adaptive notch sidelobe canceller proposed by Howells opened up a new research direction for array signal processing. Compared with the traditional single-sensor signal processing, the array signal has the advantages of strong anti-interference capability, flexible beam control, large signal gain, high spatial resolution and the like. Therefore, since the 70 s in the 20 th century, array signal processing has been widely used in civilian fields such as wireless communication, intelligent conference systems, seismic exploration, wireless astronomy, machine state monitoring and fault diagnosis, and military fields such as sonar and radar. As a core technology and method of array signal processing, a Direction of arrival (DOA) estimation method based on an array output signal feature subspace can realize super-resolution estimation of a target source Direction of arrival, and is widely used.

For the study of the super-resolution direction-of-arrival estimation problem, it is generally assumed that the array received signal is radiated from a point target source. When the radiation surface of the target is smaller than the resolution of the array, the physical size of the target source in space can be ignored and abstracted into a point in a geometric sense, so that the target source can be regarded as a point target source (referred to as a point source for short), and the point source model of the signal is shown in fig. 1. The classical direction of arrival estimation algorithm mostly assumes that line-of-sight propagation exists between the information source and the observation array and the strength of the direct wave signal is high, so that a point source model can effectively approach the actual propagation environment of the signal, and the point source is taken as a reasonable assumption, thereby greatly simplifying the complexity of DOA estimation.

However, a point source is only a mathematical assumption, which is not possible in practical applications, and the real target source radiation surface always has a certain size. In addition, with the continuous development of scientific technology, the space electromagnetic environment becomes more and more complex, and in addition to the rapid progress of human society, the number of high-rise buildings increases in geometric progression. As shown in fig. 2, in such an environment, the energy of the signal received by the sensor array exhibits a certain dispersion in a certain range of space due to multipath effects caused by scattering, reflection, diffraction, or the like during the propagation of the radio signal. The signal source is here called a distributed source (shortly: distributed source) and the signal it radiates is also called a distributed source signal. Distributed source model as shown in fig. 3, the point source signal model is not suitable for such a case, which will cause the super-resolution direction of arrival estimation algorithm based on the point source assumption to fail. For the problem of original signal diffusion caused by multipath propagation of the signal, a distributed source model usually describes the spatial distribution of a signal source by using two parameters, namely a central direction of arrival and a diffusion angle: the energy of the distributed source is concentrated in a spatial region with the central arrival direction as the center and the diffusion angle as the radius.

With the continuous progress of science and technology, the estimation method of the direction of arrival is increasingly perfected, and a large number of high-precision and high-resolution estimation methods are proposed. However, most of these methods are based on classical point source models, i.e. it is often assumed that the signal to be estimated is narrowband, and the reciprocal of the signal bandwidth is larger than the time required for the wave front to sweep through the array space. This assumption is applicable in most cases. However, in recent years, due to the increasingly complex space electromagnetic environment and the increasingly dense ground buildings, there is often no direct signal from the signal source to be estimated to the receiving sensor array, but the signal reaches the receiving array in a multipath form after being reflected and refracted. Therefore, in this case, regardless of the physical size of the target source to be estimated, abstracting it as a geometric point will reduce the accuracy of parameter estimation to a great extent, and on this basis, the concept of distributed source is proposed and successfully applied to the direction of arrival estimation.

Although the estimation of the direction of arrival and the divergence angle can be realized by using the feature subspace technology as a basic method for the distributed source condition in the array signal processing, two-dimensional spectral peak search is required, so that the calculation complexity is increased in a geometric series manner on the premise of ensuring that the parameter estimation accuracy is unchanged, the real-time performance of the algorithm is seriously influenced, and the method is greatly limited in practical application.

Disclosure of Invention

In order to overcome the problem that the estimation and calculation complexity of the direction of arrival and the diffusion angle of a target information source is higher due to the multipath effect caused by shielding, reflection and even refraction in the radio wave propagation process of the conventional direction of arrival estimation method, the invention provides a novel method for quickly estimating the direction of arrival and the diffusion angle of a multipath signal, which specifically comprises the following steps:

a distributed source model of array output signals is constructed, and the problem of low accuracy of estimation of the direction of arrival under a point source model is effectively solved;

acquiring a covariance matrix of array output signals, performing singular value decomposition on the covariance matrix to obtain a signal subspace and a noise subspace, constructing a two-dimensional space spectrum through the noise subspace, and estimating the direction of arrival and the diffusion angle of a distributed source in the two-dimensional space spectrum;

describing a deterministic angle signal density function of a distribution source by adopting a unimodal conjugate symmetric function, and reconstructing a generalized array manifold of a uniform linear array by combining a small angle approximation theory to realize the separation of the direction of arrival and the diffusion angle;

constructing a quadratic optimization function to obtain a one-dimensional space spectrum of the direction of arrival, and obtaining estimation of the direction of arrival through the one-dimensional space spectrum;

and acquiring a one-dimensional spatial spectrum of the diffusion angle, and acquiring diffusion angle estimation through the one-dimensional spatial spectrum.

Due to the adoption of the technical scheme, the invention can obtain the following technical effects: the method accurately realizes the rapid estimation of the direction of arrival and the diffusion angle of the target information source, converts the two-dimensional spectral peak search with high computation complexity into the one-dimensional spectral peak search with low complexity on the premise of ensuring the parameter estimation precision, improves the reliability of the algorithm, reduces the computation complexity of the algorithm, and can be suitable for radio monitoring and positioning scenes with higher real-time requirements.

Drawings

FIG. 1 is a diagram of a point source model;

FIG. 2 is a diagram of a distributed source model;

FIG. 3 is a diagram of electromagnetic wave multipath propagation;

FIG. 4 is a graph comparing the FLOM algorithm, PFLOM algorithm, CRCO algorithm, and DSPE algorithm of the present invention under different generalized signal-to-noise ratios;

FIG. 5 is a graph comparing the FLOM, PFLOM, CRCO, and DSPE algorithms of the present invention under different fast beat conditions.

Detailed Description

The embodiments of the present invention are implemented on the premise of the technical solution of the present invention, and detailed embodiments and specific operation procedures are given, but the scope of the present invention is not limited to the following embodiments.

Example 1

The embodiment provides a new method for quickly estimating the direction of arrival and the spread angle of a multipath signal, which may include the following steps:

firstly, constructing a distributed source model of an array output signal;

specifically, it is assumed that L distributed source signals are incident on a uniform linear array, the uniform linear array includes M signal receiving sensors, and a signal vector x output by the sensors can be expressed as:

wherein, for the ith distributed source, s_i(θ,ψ_i) In order to be the angular signal density,direction of arrival including target sourceAnd the angle of divergence sigma_iA (θ) is a guide vector of the uniform linear array, n is a noise vector, and Θ ═ π/2, π/2]Is the observation range of the distributed source.

Considering that signal and noise are uncorrelated, the correlation matrix of the signal vector x can be expressed as:

R_xx＝E(xx^H)＝R_ss(ψ)+R_nn (2)

wherein R is_ss(psi) is the correlation matrix of the clean signal, R_nnIs the correlation matrix of the noise.

The angular cross-correlation kernel is defined as:

then R is_ss(ψ) can be rewritten as:

typically, multiple target sources are uncorrelated, so the angular cross-correlation kernel can be reduced to:

ρ_ij(θ,θ′,ψ_i,ψ_j)＝δ_ij·μ_i(θ,θ′,ψ_i) (5)

wherein, delta_ijIs a Kroneckerdelta function and the angular autocorrelation kernel for the ith target source is:

at this time, the angular signal density of the distributed source is:

s_i(θ,ψ_i)＝χ(θ,ψ_i)ζ_i (7)

wherein, x (theta, psi)_i) As a function of the deterministic angular signal density, ζ_iIs a complex random variable.

Secondly, acquiring a covariance matrix of array output signals, performing singular value decomposition on the covariance matrix to obtain a signal subspace and a noise subspace, constructing a two-dimensional space spectrum through the noise subspace, and estimating a distributed source direction of arrival and a diffusion angle in the two-dimensional space spectrum, which specifically includes:

s2.1, constructing a covariance matrix of an array output signal x;

s2.2, singular value decomposition is carried out on the covariance matrix, namely M eigenvalues are sorted from large to small, eigenvectors corresponding to L large eigenvalues are divided into one group, eigenvectors corresponding to the other M-L small eigenvalues are divided into another group, and then the eigenvectors corresponding to the L large eigenvalues are the signal subspace U_sThe eigenvectors corresponding to the M-L smaller eigenvalues are the noise subspace U_n；

S2.3, constructing a two-dimensional space spectrum by using the noise subspace:

s2.4, constructing an estimation function of the arrival direction and the diffusion angle of the distributed source in the two-dimensional space spectrum:

wherein the content of the first and second substances,b (ψ) is a generalized steering vector for a uniform linear array, which is an estimate of ψ.

Describing a deterministic angle signal density function of a distribution source by adopting a single-peak conjugate symmetric function, and reconstructing a generalized array manifold of a uniform linear array by combining a small angle approximation theory to realize the separation of the direction of arrival and the diffusion angle;

specifically, the deterministic angular density function is described by a unimodal conjugate function:

at this time, in combination with the small angle approximation theory, the generalized array manifold for reconstructing a uniform linear array is:

the function h (σ) is defined as:

thus, the generalized array manifold for a uniform linear array can be further expressed as:

equation (13) achieves separation of the direction of arrival and the divergence angle.

Fourthly, constructing a quadratic optimization function to obtain a one-dimensional space spectrum of the direction of arrival, and obtaining estimation of the direction of arrival through the one-dimensional space spectrum;

specifically, a two-dimensional spatial spectrum of the distributed source is reconstructed using equation (13):

defining functionsComprises the following steps:

then

Constructing quadratic optimization functionComprises the following steps:

to ensure that the solution of the quadratic optimization function in equation (17) is not zero, the quadratic optimization function can be rewritten in the form of the following constraint:

defining functionsComprises the following steps:

solving the partial derivative with respect to h (σ):

thenThe solution of (a) is:

the one-dimensional spatial spectrum of the direction of arrival estimate is:

performing a one-dimensional spectral peak search on equation (22) may result in a direction-of-arrival estimate

Fifthly, acquiring a one-dimensional spatial spectrum of the diffusion angle, and acquiring diffusion angle estimation through the one-dimensional spatial spectrum;

specifically, willAnd (9) obtaining a one-dimensional spatial spectrum of the diffusion angle, and performing spectral peak search on the one-dimensional spatial spectrum to obtain diffusion angle estimation.

And (3) algorithm comparison: the experimental conditions set the incident angles of two distributed source target information sources respectively asAnddiffusion angles are respectively sigma₁1.5 ° and σ₂2.1. The invention is compared with FLOM algorithm, PFLOM algorithm, CRCO algorithm and DSPE algorithm. Firstly, as shown in fig. 4, as the generalized signal-to-noise ratio is increased from 0dB to 20dB, although the root mean square error and the resolvable probability of each algorithm are improved, the performance of the algorithm of the invention is always superior to that of other comparison algorithms, and two incident target information sources can be always and completely separated in the whole interval of the generalized signal-to-noise ratio change.

Secondly, as shown in fig. 5, as the number of snapshots of the array output signal samples is increased from 100 to 1000, the performance of the invention, the FLOM algorithm, the PFLOM algorithm and the CRCO algorithm is improved, but the influence of statistical errors is removed, and the performance of the DPSE algorithm is not changed significantly. However, the performance of the invention is always optimal during the variation of the number of snapshots.

The foregoing descriptions of specific exemplary embodiments of the present invention have been presented for purposes of illustration and description. It is not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and its practical application to enable one skilled in the art to make and use various exemplary embodiments of the invention and various alternatives and modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims and their equivalents.