KR101909710B1 - A method of estimating the arrival angle of the covariance matrix based on the frequency domain based on the sparsity of the signal in the sonar system and system thereof - Google Patents

Abstract

One embodiment of the present invention relates to a method of estimating a direction of arrival using a covariance matrix-based compression sensing in a frequency domain which estimates the direction of arrival of a signal based on the sparsity of a signal in a sonar system. According to the present invention, the method of estimating a direction of arrival of a covariance matrix-based compression sensing in a frequency domain comprises a block division step of performing a first transformation of a signal reflected from a target received by a sensor array into an incident signal matrix of time domain and dividing the first transformed incident signal matrix by a predetermined number of blocks; a bin selection step of performing a second transformation to process the divided receiving signal matrix in a frequency domain and selecting at least one or more beans based on the size of a bin in the result of the second transformation; a covariance matrix generation step of generating a bin vector based on the selected bin and generating a covariance matrix based on the generated bin vector; and a direction-of-arrival estimation step of applying an SpSF (Sparse Spectrum Fitting) algorithm to the generated covariance matrix and estimating the direction of arrival of the incident signal.

Description

Technical Field [1] The present invention relates to a covariance matrix-based estimation method and system for estimating an arrival angle of a covariance matrix in a frequency domain based on the scarcity of a signal in a sonar system sonar system and system thereof}

The present invention relates to a method and system for estimating an arrival angle based on a covariance matrix in a frequency domain on the basis of the scarcity of a signal in a sonar system. More particularly, the present invention relates to a method and system for estimating a covariance matrix- The present invention relates to a method and a system for extending a sensing arrival angle estimation method to a frequency domain.

Based on the compression sensing, the angular estimation method is a method of estimating the arrival angle using the sparsity of the signal, and has many advantages in comparison with the conventional arrival angle estimation algorithm and has been studied steadily. Specifically, the compression sensing technique is a technique for estimating the DOA (Direction Of Arrival) using the characteristics of the signal in the sonar system itself. Although the accuracy of the arrival angle estimation can be improved by using the signal scarcity, There are many limitations.

Also, as an example of the estimation method based on the compression sensing, the arrival angle estimation method using the covariance matrix fitting estimates the arrival angle of the incident signal by scarcely optimizing the signal covariance of the received signal through the cost function. The SpSF (Sparse Spectrum Fitting) algorithm is a typical algorithm used in this technique. Since all studies on the existing SpSF algorithm are implemented only in the time domain, it is difficult to apply them in a situation where various noises such as underwater environments exist There is a dot.

Korean Patent No. 10-1783777 (published on September 26, 2017)

SUMMARY OF THE INVENTION It is an object of the present invention to provide a covariance matrix-based arrival angle estimation method in a frequency domain rather than a time domain.

According to an aspect of the present invention, there is provided a covariance matrix-based compression sensing angle-of-arrival estimating method in a frequency domain, comprising: estimating an arrival angle of a signal based on a sparsity of a signal in a sonar system; The method comprising: a first step of first converting a signal reflected from a target received by a sensor array into a reception signal matrix in a time domain; Dividing a first transformed received signal matrix by a predetermined number of blocks; A bin selection step of performing a second transformation to process the divided reception signal matrix in the frequency domain and selecting at least one bin based on the size of the bin in the second transformed result; A covariance matrix generating step of generating a bin vector based on the selected bin and generating a covariance matrix based on the generated bin vector; And an arrival angle estimating step of estimating an arrival angle of the received signal by applying an SpSF (Sparse Spectrum Fitting) algorithm to the generated covariance matrix.

According to another aspect of the present invention, there is provided a covariance matrix-based compression sensing angle-of-arrival estimating system in a frequency domain, which estimates the arrival of a signal based on a sparsity of a signal in a sonar system, In a covariance matrix-based compression sensing arrival angle estimation system in a frequency domain for estimating angles, a sensor array converts a signal reflected from a target into a time-domain received signal matrix, A block dividing unit dividing the first transformed received signal matrix by a predetermined number of blocks; A bin selection unit that performs a second transformation to process the divided reception signal matrix in a frequency domain and selects at least one bin based on a size of a bin in the second transformed result; A covariance matrix generator for generating a bin vector based on the selected bin and generating a covariance matrix based on the generated bin vector; And an adverberating smoothing unit for applying an SpSF (Sparse Spectrum Fitting) algorithm to the generated covariance matrix to estimate an arrival angle of the received signal.

An embodiment of the present invention can provide a computer-readable recording medium storing a program for executing the covariance matrix-based compression sensing arrival-angle estimation method in the frequency domain.

According to the present invention, by generating a covariance matrix in the frequency domain instead of the time domain on the basis of the scarcity of the signal in the sonar system, and estimating the incident angle of the incident signal on the basis of the generated covariance matrix, It is possible to estimate the angle of incidence accurately compared with the arrival angle estimation method.

Further, according to the present invention, the improvement of the signal-to-noise ratio is superior to that of obtaining a covariance matrix in the time domain by estimating the covariance matrix by taking only a bin corresponding to the narrowband frequency of the incident signal in the frequency domain, It is possible to improve the estimation accuracy of the incident angle of the incident signal in an environment where it is difficult to estimate the arrival angle due to various noise such as the environment.

FIG. 1 is a diagram showing an example of a result obtained by transforming a received signal in order to process a received signal in a frequency domain, according to an arrival angle estimation system according to the present invention.
2 is a block diagram of an arrival angle estimation system according to an embodiment of the present invention.
3 is a diagram schematically illustrating an example of the arrival angle estimation result according to the present invention.
4 is a diagram schematically illustrating another example of the arrival angle estimation result according to the present invention.
5 is a diagram schematically showing another example of the arrival angle estimation result according to the present invention.
6 is a diagram schematically illustrating another example of the arrival angle estimation result according to the present invention.
FIG. 7 is a flowchart illustrating an example of a method of estimating an arrival angle based on a covariance matrix in a frequency domain according to the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. The effects and features of the present invention and methods of achieving them will be apparent with reference to the embodiments described in detail below with reference to the drawings. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings, wherein like reference numerals refer to like or corresponding components throughout the drawings, and a duplicate description thereof will be omitted .

In the following embodiments, the terms first, second, and the like are used for the purpose of distinguishing one element from another element, not the limitative meaning.

In the following examples, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise.

In the following embodiments, terms such as inclusive or possessed mean that a feature or element described in the specification is present, and does not exclude the possibility that one or more other features or components are added in advance.

If certain embodiments are otherwise feasible, the particular process sequence may be performed differently from the sequence described. For example, two processes that are described in succession may be performed substantially concurrently, and may be performed in the reverse order of the order described.

FIG. 1 is a diagram showing an example of a result obtained by transforming a received signal in order to process a received signal in a frequency domain, according to an arrival angle estimation system according to the present invention.

Referring to FIG. 1, the arrival angle estimation system according to the present invention receives a signal, firstly converts the signal into a received signal matrix, and secondarily converts the received signal matrix to process in the frequency domain.

The received signal matrix that has been transformed in order to be processed in the frequency domain is divided into various frequency components having various sizes as shown in FIG. 1. Hereinafter, as shown in FIG. 1, the received signal matrix is processed in the frequency domain Each frequency component having each magnitude calculated as a result of the conversion to be referred to as a bin will be referred to as a bin.

In FIG. 1, the (i-1) -th bin and the (i) -th bin are adjacent to each other, and all of them are found to exceed the threshold value. 1, the (i + 3) th bin and (i + 4) th bin are adjacent to each other, but the (i + 4) th bin has a size exceeding the threshold value, . ≪ / RTI > In FIG. 1, whether or not the size of each bin exceeds a threshold value and the meaning that adjacent bins are adjacent to each other will be further described with reference to FIG.

2 is a block diagram of an arrival angle estimation system according to an embodiment of the present invention.

2, the covariance matrix-based arrival angle estimation system 20 in the frequency domain according to an embodiment of the present invention includes a block division unit 210, a bin selection unit 230, a covariance matrix generation unit 250 ) And an incoming square portion 270. [0064] Hereinafter, the covariance matrix-based arrival angle estimation system 20 in the frequency domain according to the present invention will be abbreviated as the arrival-angle estimation system 20.

The compression sensing technique according to the present invention approaches a DOA estimation technique using a property that a signal is sparse in a sonar system. According to the present invention, by using the sparsity of a signal, it is possible to technically overcome the difficulty in estimating the arrival angle when a signal exceeding the limit of the number of sensor elements is incident.

The block dividing unit 210 receives the signal received by the linear array sensor, converts the received signal matrix into a time-domain received signal matrix, and divides the received signal matrix by a predetermined number of blocks. The signal received by the block division unit 210 is a signal received by a sensor array, which means a signal reflected from a target.

In this case, a block is an abstract concept for uniformly dividing a received signal matrix, and the number of blocks for dividing the received signal matrix as a constant value or a logic value may be preset in the block dividing unit 210.

As an example of the number of constant value blocks set in the block dividing unit 210, the number of blocks may be 2000. As another example, the number of blocks may be equal to the number of sensors used to sense the received signal, or the number of sensors May be any arbitrary number. For convenience of explanation, the number of blocks will hereinafter be referred to as " L ".

In an alternative embodiment, the received signal may include signal size information, frequency information of the signal, and sensor number information on the number of sensors used to sense the signal, and the block splitting unit may include a transformed received signal matrix May be equal to or greater than the number of sensors included in the sensor number information. According to this alternative embodiment, it becomes possible to estimate a more accurate incident angle at a higher speed as it is divided into blocks sufficiently to accurately estimate the incident angle of the incident signal.

The bin selection unit 230 performs a transform for processing the received signal matrix in the frequency domain, and selects at least one bin based on the size of the bin in the transformed result. Hereinafter, in order to clearly distinguish that the bin selection unit 230 transforms the received signal matrix, and that the block division unit 210 transforms the received signal into the received signal matrix, .

As an alternative embodiment in which the bin selection unit 230 performs a second transformation on the received signal matrix, the bin selection unit 230 may perform a second transformation on the received signal matrix through Fast Fourier Transform.

As an alternative embodiment of the method in which the bin selecting unit 230 selects a bin from the second transformed result, the bin selecting unit 230 may determine whether the size of the bin and the size of adjacent bins are larger than the size of the adjacent bin As a basis, at least one bin may be selected. The bin selection unit 230 does not use only one bin in which the signal energy exists for accurate arrival angle estimation. The bin selection unit 230 takes the largest size bin in the second conversion result and generates a component covariance matrix of adjacent bin when the signal size ratio of the bin and the signal size ratio of the adjacent bin exceeds the threshold value To be used. 1, the bin selection unit 230 determines that the (i + 3) -th bin and the (i + 4) -th bin do not satisfy the threshold value, The beans are not selected to be used to generate the covariance matrix. As another example, the bin selection unit 230 may select to use the i-th bin and the i-th bin in Fig. 1 for generating a covariance matrix, based on exceeding the threshold value.

For example, the closer the frequency of the tone is to the frequency at which the frequency bin exists, the smaller the ratio of the maximum signal size to the adjacent signal size, and the beacon selector 230 at this time takes only the beans having the maximum signal size .

As another example, as the frequency of the tone moves away from the frequency at which the frequency bin exists, the ratio of the magnitude of the maximum signal to the magnitude of the adjacent signal increases. In this case, Takes a component of the bin and generates a covariance matrix.

As another example, when the frequency of the tone is close to an integral multiple of fs / N, energy is concentrated at one frequency, so that the ratio of the signal of the largest bin to the signal of the adjacent bin decreases, 230) will only select the largest bean.

In another example, if the frequency of the tone is close to (integer + 0.5) times of fs / N, energy is scattered in the adjacent two bins, so that the ratio of the largest signal to the adjacent bin signal increases, In this case, the bin selector 230 selects a plurality of bins so that a matrix in which not only the largest bin but also a neighboring bin can be generated.

The covariance matrix generator 250 generates an empty vector based on the bin selected by the bin selector 230, and generates a covariance matrix based on the generated empty vector.

로 정의할 수 있다.When the bin in which the signal energy is present is the i-th bin, the bin adjacent to the i-th bin, and the bin with the signal size ratio equal to or greater than the threshold value is the (i-1) -th bin, the covariance matrix generating unit 250 generates Made bin vector

.

Equations (1) to (7) collectively show the mathematical equations necessary for generating the covariance matrix of the signal received by the covariance matrix generator 250 when the received signals (incident signals) have a correlation .

Equation (1) represents an empty vector composed of i-1 < th > bin and i < th > bin.

는 표적신호벡터

와 잡음신호벡터

로 구성되는 것을 알 수 있다.Equation (2) and Equation (3) represent the received signal vector, which means that the received signal vector is expressed as a vector of the received signal matrix. According to Equation (2), the received signal vector

Lt; RTI ID = 0.0 >

And the noise signal vector

. ≪ / RTI >

Equation (4) represents a target signal vector, Equation (5) represents a noise (noise) vector, and Equations (6) and (7) represent a coefficient vector multiplied to a target signal vector.

는 본 발명에 따른 도래각 추정 시스템(20)이 탐색하고자 하는 도래각의 개수이다.In Equations 1 to 7, M is the number of sensors used to sense the received signal,

Is the number of arrival angles to be sought by the arrival angle estimation system 20 according to the present invention.

The covariance matrix generator 250 may obtain the covariance matrix obtained from the block averages in the frequency domain by using Equation (8) after receiving the information according to Equations (1) to (7) from the bin selector 230 .

Equation (8) is an example of a formula used by the covariance matrix generator 250 to obtain the covariance matrix of the received signal.

는 j번째 블록의 빈벡터를 의미하고, L은 블록의 개수,

는 수신된 신호(입사신호)의 공분산행렬, E는 잡음 공분산행렬을 의미한다. In Equation (8)

Denotes an empty vector of the jth block, L denotes the number of blocks,

Denotes a covariance matrix of the received signal (incident signal), and E denotes a noise covariance matrix.

The covariance matrix generator 250 can generate the signal covariance matrix using Equation (8) by separating the incident signal when the incident signal has no correlation with the normal correlation.

In the frequency domain, the arrival angle estimation method using the SpSF algorithm is divided into the arrival angle estimation of the uncorrelated signal and the arrival angle estimation of the correlated signal having the general correlation according to the incident signal correlation.

를 구하기 위해 사용하는 비용함수는 수학식 9와 같다.First, when the incident signal has a general correlation, the covariance matrix of the signal has a value other than 0 in both the diagonal component and the other components, so that the covariance matrix generator 250 generates the most sparse signal covariance matrix

(9) " (9) "

행렬이고, 탐색하는 도래각의 개수인

는 센서의 배열개수인 M보다 훨씬 더 많기 때문에 수학식 1은 비결정시스템(underdetermined system)이고, 수학식 8을 만족하는 해는 무수히 많이 존재한다.Equation (9) represents a cost function used by the covariance matrix generator 250 to obtain the most rare signal covariance. The array matrix A

Matrix, and the number of incident angles to be searched

(1) is an underdetermined system, since there are much more than M, which is the number of the array of sensors, and there are a large number of solutions satisfying the expression (8).

와 A의 곱을 통해 산출되는 공분산벡터의 차이(difference)를 나타낸다. 보다 구체적으로, L2-norm은 최적화하는 과정에서 추정된 값들 중 참값과 가장 비슷한 추정값을 선택하는 부분이다.In Equation (9), the L2-norm (first term) operation portion includes a covariance vector R and a signal covariance vector

And the difference between the covariance vectors calculated through the product of A and A. More specifically, the L2-norm is a portion that selects the estimation value that is closest to the true value among the estimated values in the optimization process.

의 희소성(sparsity)을 조절하는 부분이다. 가중치 λ값을 크게 설정함으로써, 기존의 신호와 연산을 통해 구한 추정데이터와의 차이를 줄이는 것보다 추정 데이터의 희소성을 강조한다. 그와 반대로, 가중치 λ값을 작게 설정하게 되면, 데이터의 희소성보다는 기존의 신호와의 차이를 최소화하는 것을 강조한다. 수학식 9와 같은 비용함수를 통해서, 최적화 작업을 반복함으로써, 공분산행렬생성부(250)는 가장 희소한 신호공분산 벡터를 구할 수 있게 된다.Then, the L1-norm (second term) operation portion of Equation (9)

Which is the part that controls the sparsity. By setting the weighting lambda value large, the scarcity of the estimated data is emphasized rather than reducing the difference from the estimated data obtained through calculation with the existing signal. Conversely, by setting the weighting lambda value small, it is emphasized that the difference from the existing signal is minimized rather than the scarcity of the data. By repeating the optimization operation through the cost function shown in Equation (9), the covariance matrix generator 250 can obtain the most rare signal covariance vector.

를 정의하게 된다.The covariance matrix generator 250 has a property that when the incident signals are not correlated, the covariance matrix of the received signal has a value other than zero for the diagonal components and all values other than the diagonal components are zero. Therefore, the covariance matrix obtained by the block average in the frequency domain can be expressed by Equation (10). The signal covariance of the uncorrelated signal is also a nonzero value. In order to vectorize the signal covariance, the covariance matrix generator 250 calculates the signal covariance vector

.

(10) to (15) collectively indicate mathematical equations necessary for generating a covariance matrix of the signal received by the covariance matrix generator 250 when the received signals (incident signals) do not correlate .

Equations (10) to (15) are mathematical expressions corresponding to Equations (2) to (7), and show a received signal matrix (vector) calculated from an incident signal. The SpSP cost function when the received signal is not correlated is expressed by Equation (16).

를 구할 수 있다. 공분산행렬생성부(250)는 SpSF 비용함수에서 가중치 λ의 값을 정하는 방법인 경우, λ의 값을 양의 정수 방향으로 바꾸어가면서 신호공분산 행렬

(uncorrelated 신호인 경우에는

)을 구하는 과정을 반복하고, 위와 같은 반복과정을 통해 생성된 신호공분산 행렬들 중에서 수학식 17을 만족하는 λ값을 가중치로 결정한다.Equation 16 represents a cost function used by the covariance matrix generator 250 to generate the signal covariance matrix when the received signals have no correlation. The covariance matrix generator 250 may adjust the value of the weighting factor λ in Equation 16 as described in Equation 9 so that the most similar signal covariance matrix most similar to the covariance of the received signal

Can be obtained. The covariance matrix generator 250 converts the value of [lambda] into the direction of a positive integer and determines a signal covariance matrix

(if it is an uncorrelated signal

), And determines a value of lambda that satisfies Equation (17) as a weight among the signal covariance matrices generated through the above-described iterative process.

Equation (17) represents a mathematical expression used by the covariance matrix generator 250 to obtain the weighting factor? When the received signal is not correlated.

를 수학식 17과 같이 유도함으로써, 수학식 9와 수학식 16을 통해서 가장 적합한 가중치 λ를 구할 수 있다.Since the covariance matrices of noise are practically not available at the time of signal processing, the statistical properties of the noise covariance matrices are used

Is derived as shown in Equation (17), then the most suitable weighting value? Can be obtained through Equations (9) and (16).

The covariance matrix generator 250 may previously store a correlation reference value for determining the correlation of the received signal.

The arrival quadrature determiner 270 estimates the arrival angle of the received signal by applying the SpSF algorithm to the covariance matrix generated by the covariance matrix generator 250. [ More specifically, the arrival quadrature determiner 270 performs an SpSF algorithm in the frequency domain on the basis of the signal covariance matrix, performs a spectrum of the diagonal components of the signal covariance matrix, And a method of estimating the arrival angle of a received signal by applying the SpSF algorithm to a previously determined covariance matrix is well known, and thus a detailed description thereof will be omitted.

3 is a diagram schematically illustrating an example of the arrival angle estimation result according to the present invention.

More specifically, FIG. 3 shows the case where the number of incident signals is two (30 degrees, 75 degrees), the number of sensors is 10, the number of blocks is 2000, the signal frequency is equal to 400 Hz, and the signal- This is the estimation result of the arrival angle.

Referring to FIG. 3, a signal covariance matrix is generated by taking a bin implemented in the present invention when a signal-to-noise ratio is 0 dB and two signals are incident. In the case of the environment, the SNR is low, It can be seen that the power is high. The covariance matrix generation unit 250 generates a covariance matrix by taking three beans in total, and applies it to the SpSF algorithm (CS algorithm). In Fig. 3, according to the MVDR and SpSF algorithm, the largest spectral value is obtained at the position where two signals are incident. Also, in FIG. 3, according to the CBF algorithm, it can be seen that an accurate estimation is performed for only one of the incident signals and an estimation error occurs for the remaining signals.

4 is a diagram schematically illustrating another example of the arrival angle estimation result according to the present invention.

More specifically, FIG. 4 shows a case where the number of incident signals is two (15 degrees, 30 degrees), the number of sensors is 10, the number of blocks is 2000, the signal frequency is 400 Hz and the signal- This is the estimation result of the arrival angle.

FIG. 4 shows the results of the arrival angle estimation of each algorithm when the incident angle is close to 15 degrees and 30 degrees when the signal-to-noise ratio is 10 dB and two signals are incident. In FIG. 4, the covariance matrix generator 250 uses only one bin when generating a covariance matrix considering that the signal-to-noise ratio is 10 dB and adjacent bins have a small power value less than the threshold value. 4, only the SpSF algorithm (CS algorithm) extended to the frequency domain accurately estimates the angle of incidence by correctly decomposing the two angles of the incident signal, whereas the CBF or MVDR algorithm does not accurately estimate the angle of incidence. .

5 is a diagram schematically showing another example of the arrival angle estimation result according to the present invention.

More specifically, FIG. 5 shows a case where the incident signal is two (15 degrees, 30 degrees), the number of sensors is 10, the number of blocks is 2000, the signal frequency is 400 degrees, 380 Hz, and the signal-to-noise ratio is 10 dB, it is the arrival angle estimation result in the case where the arrival angle of the incident signal is 15 degrees.

5, the covariance matrix generator 250 has a signal-to-noise ratio of 10 dB, so that the bins adjacent to the bin having the largest power have small power, so that only one bin is used to generate the covariance matrix, 5, it can be seen that both the CBF, MVDR and SpSF algorithms can normally estimate the incident angle of the incident signal.

6 is a diagram schematically illustrating another example of the arrival angle estimation result according to the present invention.

More specifically, Fig. 6 shows a case where the incident signal is two (15 degrees, 30 degrees), the number of sensors is 10, the number of blocks is 2000, the frequency is 400 Hz when the signal frequency is 15 degrees, 380 Hz, a signal-to-noise ratio of 10 dB, and an incident angle of the incident signal of 30 degrees.

6, the covariance matrix generator 250 has a signal-to-noise ratio of 10 dB, so that the bins adjacent to the bin having the largest power have a small power, so that only one bin is used to generate the covariance matrix, 6, it can be seen that both the CBF, MVDR and SpSF algorithms can normally estimate the incident angle of the incident signal.

5 and 6, when a covariance matrix is generated according to the present invention when the incident angles of the incident signals are different from each other and the frequencies are different from each other, when the covariance matrix is applied to all the conventional estimation algorithms, Can be estimated normally, but it can be seen from the spectrum that SpSF performance is better than CBF or MVDR algorithm.

3 and 4, if the signal-to-noise ratio is higher than a certain level when the incident angles of the incident signals are different from each other and the frequencies are equal to each other, a covariance matrix is generated according to the present invention, It is shown that the best performance is obtained by applying the SpSF algorithm based on the covariance matrix and estimating the angle of incidence of the incident signal.

FIG. 7 is a flowchart illustrating an example of a method of estimating an arrival angle based on a covariance matrix in a frequency domain according to the present invention.

Since the method according to FIG. 7 can be implemented by the arrival angle estimation system 20 described with reference to FIG. 2, a description overlapping with the description with reference to FIG. 2 will be omitted. Hereinafter, do.

The block dividing unit 210 first converts the received signal into a received signal matrix, and divides the received signal matrix by a predetermined number of blocks (S710).

The bin selection unit 230 performs a second transformation to process the received signal matrix in the frequency domain, and selects at least one bin based on the size of the bin in the second transformed result (S730).

The covariance matrix generator 250 generates an empty vector based on the bin selected by the bin selector 230, and generates a covariance matrix using the generated empty vector (S750).

The arrival quadrature determiner 270 estimates the arrival angle of the received signal by applying the SpSF algorithm to the covariance matrix generated by the covariance matrix generator 250 (S770).

If the incident signal is a narrowband signal in the time domain, the time domain signal has a large value for all values of the time domain index. On the other hand, due to the characteristics of the narrowband signal, when the time domain signal is converted into the frequency domain signal through the FFT (DFT), the narrowband signal has a value only in certain frequency bins and has a very small value in other part bins, , The frequency domain spectrum of the noise has almost the same value for all bins.

Therefore, estimating the covariance matrix by taking only the bin corresponding to the narrowband frequency of the incident signal in the frequency domain has an effect of improving the signal-to-noise ratio compared to obtaining the covariance matrix in the time domain. This phenomenon is maximized when a sinusoidal wave (sinusoidal signal, or tone signal), which is dense and all energy is dense at one frequency, is received as an incident signal. The present invention has the effect of improving the arrival angle estimation performance of an incident signal in a situation where it is difficult to estimate an arrival angle due to various noises such as an underwater environment due to the above characteristics.

The embodiments of the present invention described above can be embodied in the form of a computer program that can be executed on various components on a computer, and the computer program can be recorded on a computer-readable medium. At this time, the medium may be a magnetic medium such as a hard disk, a floppy disk and a magnetic tape, an optical recording medium such as CD-ROM and DVD, a magneto-optical medium such as a floptical disk, , A RAM, a flash memory, and the like, which are specifically configured to store and execute program instructions.

Meanwhile, the computer program may be designed and configured specifically for the present invention or may be known and used by those skilled in the computer software field. Examples of computer programs may include machine language code such as those produced by a compiler, as well as high-level language code that may be executed by a computer using an interpreter or the like.

The specific acts described in the present invention are, by way of example, not intended to limit the scope of the invention in any way. For brevity of description, descriptions of conventional electronic configurations, control systems, software, and other functional aspects of such systems may be omitted. Also, the connections or connecting members of the lines between the components shown in the figures are illustrative of functional connections and / or physical or circuit connections, which may be replaced or additionally provided by a variety of functional connections, physical Connection, or circuit connections. Also, unless explicitly mentioned, such as " essential ", " importantly ", etc., it may not be a necessary component for application of the present invention.

The use of the terms " above " and similar indication words in the specification of the present invention (particularly in the claims) may refer to both singular and plural. In addition, in the present invention, when a range is described, it includes the invention to which the individual values belonging to the above range are applied (unless there is contradiction thereto), and each individual value constituting the above range is described in the detailed description of the invention The same. Finally, the steps may be performed in any suitable order, unless explicitly stated or contrary to the description of the steps constituting the method according to the invention. The present invention is not necessarily limited to the order of description of the above steps. The use of all examples or exemplary language (e.g., etc.) in this invention is for the purpose of describing the present invention only in detail and is not to be limited by the scope of the claims, It is not. It will also be appreciated by those skilled in the art that various modifications, combinations, and alterations may be made depending on design criteria and factors within the scope of the appended claims or equivalents thereof.

Claims (9)

A covariance matrix-based compression sensing arrival angle estimation method in a frequency domain based on the sparsity of a signal in a sonar system,
A sensor array first converts a signal reflected from a target into a time-domain received signal matrix, and divides the first received signal matrix by a predetermined number of blocks Block dividing step;
Calculating a plurality of bins by performing a second transformation to process the divided reception signal matrices in a frequency domain and calculating at least one of the plurality of bins based on the size of the calculated bins and whether or not the size of the adjacent bins exceeds a size reference An empty selection step of selecting more than one bin;
A covariance matrix generating step of generating a bin vector based on the selected bin and generating a covariance matrix based on the generated bin vector; And
And an arrival angle estimating step of estimating an arrival angle of the received signal by applying SpSF (Sparse Spectrum Fitting) algorithm to the generated covariance matrix. The method according to claim 1,
Wherein the second conversion comprises:
Wherein the fast Fourier transform is a Fast Fourier Transform (FFT). The method according to claim 1,
The received signal may be transmitted,
The signal size information, the frequency information of the signal, and the sensor number information on the number of sensors used to sense the signal,
Wherein, in the block dividing step,
Wherein the first transformed received signal matrix is divided by the number of blocks equal to or greater than the number of sensors included in the sensor number information or the number of blocks larger than the number of the sensors. Respectively. The method according to claim 1,
The received signal may be transmitted,
And information on the correlation of the signal,
The covariance matrix generating step may include:
A covariance matrix based on a covariance matrix in a frequency domain is generated by dividing a predetermined correlation criterion by a case where the correlation is not correlated with a correlation based on the received signal, Way. A computer-readable recording medium storing a program for executing the method according to any one of claims 1 to 4. In a covariance matrix-based compression sensing arriving angular estimation system in the frequency domain based on the sparsity of a signal in a sonar system,
A sensor array first converts a signal reflected from a target into a time-domain received signal matrix, and divides the first received signal matrix by a predetermined number of blocks Block dividers;
A bin selection unit that performs a second transformation to process the divided reception signal matrix in a frequency domain and selects at least one bin based on a size of a bin in the second transformed result;
A covariance matrix generator for generating a bin vector based on the selected bin and generating a covariance matrix based on the generated bin vector; And
And estimating an arrival angle of the received signal by applying an SpSF (Sparse Spectrum Fitting) algorithm to the generated covariance matrix based on the covariance matrix based on the covariance matrix. The method according to claim 6,
Wherein the second conversion comprises:
Wherein the fast Fourier transform is a fast Fourier transform (FFT). The method according to claim 6,
The received signal may be transmitted,
The signal size information, the frequency information of the signal, and the sensor number information on the number of sensors used to sense the signal,
Wherein the block division unit comprises:
Wherein the first transformed received signal matrix is divided by the number of blocks equal to or greater than the number of sensors included in the sensor number information or the number of blocks larger than the number of the sensors. Each estimation system. The method according to claim 6,
The received signal may be transmitted,
And information on the correlation of the signal,
Wherein the covariance matrix generator comprises:
Wherein a covariance matrix is generated in a different manner by distinguishing a case in which the received signal has no correlation and a case in which the received signal has no correlation based on a predetermined correlation criterion, system.

Images