JP2002107440A - Method and device for detecting direction - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a direction detecting method and a direction detecting device capable of accurately detecting the arrival direction of an arrival wave with excellent resolution even if a spatial spreading is generated in the arrival wave. SOLUTION: This is a method for detecting the arrival direction of signal by receiving one or more signals arriving from any direction with an array antenna, and detecting the arrival direction of each received signal. This method uses a function or an equation including steering vector a (θ0-jΔθ) using the center θ0 in the arrival direction of each received signal and the spreading Δθin the arrival direction, and computes the function or the equation on the basis of the signal received by each element of the array antenna so as to obtain the center θ0 and the spreading Δθ of the arrival direction of each signal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a direction detecting method and a direction detecting apparatus for receiving a plurality of waves arriving from mutually different directions by an array antenna and detecting the direction of arrival of each wave.

[0002]

2. Description of the Related Art As a method for receiving waves arriving from different directions by an array antenna and detecting the direction of arrival of each wave, an algorithm conventionally called "MUSIC" (RO Schmidt, "Multiple Emitter Lo
cation and Signal Parameter Estimation ”, IEEE Tran
s, Antenna, Propagat., AP-34, No.3, pp.276-280, Ma
r. 1986.) and an algorithm called “ESPRIT” (R. Roy, T. Kailath, “ESPRIT-Estimation of Sign
al Parameters via Rotational Invariance Technique
s ”, IEEE Trans. Acoust., Speech, Signal Processin
g, ASSP-37, No. 7, pp. 984-995, July 1989.).

[0003] These algorithms are widely used in direction finding devices because they can achieve better resolution and accuracy than methods using Fourier transform and methods using beam scanning.

[0004]

However, the conventional "MUSIC" and "ESPRIT" algorithms are
It is based on a model that assumes that the wave arriving at the array antenna is a plane wave. Therefore, basically, when a wave having a spread arrives, its arrival direction cannot be estimated.

For example, assuming that radio waves radiated from a mobile station are incident on an antenna of a base station in a mobile communication environment, radio waves radiated from the mobile station may cause obstacles such as buildings existing around the mobile station. After being scattered under the influence of an object, the light enters the base station. In this case, it is more appropriate to consider that the incoming wave entering the base station arrives with a spatial spread. That is, it can be considered that the actual arrival direction of the incoming wave comes from not only the direction of the mobile station but also various directions near the direction of the mobile station.

Here, a signal (narrowband signal) S _i (t) (where i = 1, 2,...) From each of the d mobile stations.
d) if arrives, a case of receiving by using an array antenna of m elements which are arranged at equal intervals on a straight line at intervals d _x. In this case, an output vector y (t) representing the reception output of the array antenna is represented by the following equation.

(Equation 1) For example, when the incoming wave incident on the array antenna of the base station does not show a spatial spread, as in the case where the space between the mobile station and the base station can be seen through, the space of the formula (1) is used. response v _i is expressed by the following equation.

(Equation 2) Here, θ _i is the arrival direction of the signal from the i-th mobile station, and represents an angle from a direction perpendicular to the array antenna.
A (θ _i ) is an element corresponding to the direction of θ _i and is called a maneuver vector. The steering vector a (θ _i ) is represented by the following equation.

(Equation 3) In the equation (3), the phase difference between adjacent terms (exp (-jkd _x si
n (θ _i ))) is called a mode for the angle of arrival θ _i . In the “MUSIC” and “ESPRIT” algorithms, the arrival angle is estimated by obtaining this mode.

However, as described above, in an environment in which an incoming wave incident on an array antenna of a base station is spatially spread, the above-described mode cannot express a spatial response. That is, when the incoming wave has a spatial spread, the conventional "MUSIC" or "ESPRI"
The direction of arrival of an incoming wave cannot be estimated using the algorithm of "T".

SUMMARY OF THE INVENTION It is an object of the present invention to provide a direction detecting method and a direction detecting apparatus capable of detecting an incoming direction with excellent resolution and accuracy even when an incoming wave has a spatial spread. I do.

[0009]

A first aspect of the present invention is a direction detecting method for receiving one or a plurality of signals arriving from arbitrary directions by an array antenna and detecting a direction in which each of the received signals arrives. Then, using a function or equation including a maneuvering vector a (θo-jΔθ) with the center of the arrival direction θo and the spread of the arrival direction Δθ of each of the received signals as variables, using each element of the array antenna The above function or equation is calculated based on the received signals, and the center θo and the spread Δθ of the arrival direction of each signal are obtained.

A leaky wave is known as a mode representing the spatial spread of a wave (RE Collin, FJ Zucker, “Ante
nna Theory Part 2 ”, Chapter 20, McGraw-Hill, 196
9.). This leaky wave mode is generally expressed by the following equation using a phase rotation and an attenuation mode.

(Equation 4) This leaky wave mode can be considered to be the same as the radiation pattern of an array antenna densely arranged on the x-axis. The radiation pattern R (θ) of this array antenna is expressed by the following equation.

(Equation 5) Further, assuming that the attenuation amount α is much smaller than 1, the direction θo of the peak (center) of the radiation pattern and its spread width 2Δθ (the angle within a range where the attenuation from the peak point is within 3 dB) are expressed by the following equations. It is represented by

(Equation 6) By using these variables (θo, Δθ), the leaky wave mode of the above equation (4) can be expressed by the following equation.

(Equation 7) The right side of the above equation (8) represents the mode of the arriving wave whose center direction is θo and whose 3dB spread width is 2Δθ. Therefore, spatial response v _i of i-th arrival wave having a spatial extent it can be expressed by the following equation.

(Equation 8) That is, by finding a mode corresponding to the angle of the incoming wave represented by a complex number, the center θ of the arrival direction of each incoming wave is obtained.
o and 3 dB spread width 2Δθ can be obtained. Therefore, in claim 1, a function or an equation including a steering vector a (θo−jΔθ) having θo and Δθ as variables is used, and the function or the equation is calculated based on a signal received by each element of the array antenna. Is calculated. Therefore, the center θo and the spread Δθ of the arrival direction of each arriving wave
Can be obtained with high accuracy.

A second aspect of the present invention is a direction detecting apparatus for receiving one or a plurality of signals arriving from arbitrary directions by an array antenna and detecting a direction in which each of the received signals arrives, Sampling means for sampling the reception waves received by the respective elements at predetermined time intervals to obtain a predetermined number of signal values as a first matrix, and a sampling means for forming a covariance matrix from the first matrix. A variance matrix creation unit, an eigenvalue decomposition unit that extracts a plurality of eigenvalues and eigenvectors by decomposing the covariance matrix created by the covariance matrix creation unit, and the array antenna based on the eigenvalues extracted by the eigenvalue decomposition unit. Signal number detecting means for detecting the number of arriving signals, and a maneuvering vector a (θo jΔθ) and a direction calculation means for obtaining θo and Δθ necessary to maximize the function by using a function including the noise component En of the eigenvector as an element and the eigenvector of the noise component of the received wave. It is characterized by the following.

According to a second aspect of the present invention, the direction calculating means comprises θo and Δθ
Is used to minimize the function by using a function including the maneuvering vector a (θo-jΔθ) and the eigenvector noise component En as elements and the eigenvector of the received wave noise component as variables. Find Δθ. That is,
By using an algorithm obtained by modifying the conventional “MUSIC” algorithm using the maneuvering vector a (θo−jΔθ) of the equation (9), the direction of the arriving wave
o and the spread width 2Δθ can be obtained with high accuracy.

The eigenvalues extracted by the eigenvalue decomposition means correspond to each component of the incoming wave. In addition, an eigenvector is obtained for each eigenvalue. The eigenvector is a weight vector used for extracting each arriving wave component from the received signal. According to a third aspect of the present invention, in the direction detecting apparatus of the second aspect, spectrum drawing means for displaying a change in the value of the function according to each change of the θo and Δθ in a two-dimensional space is provided as the direction calculation means. It is characterized by the following.

According to a third aspect of the present invention, by referring to a change in the value of the function displayed by the spectrum drawing means, a position where the value of the function is minimized is visually examined to determine the direction θo and the spread 2Δθ of the arriving wave. Can be detected. Claim 4 is a direction detecting device that receives one or a plurality of signals respectively arriving from arbitrary directions by an array antenna and detects a direction in which each received signal arrives, wherein each of the array antennas Sampling means for sampling the reception waves received by the elements at predetermined time intervals to obtain a predetermined number of signal values as a first matrix, and covariance matrix creation for creating a covariance matrix from the first matrix Means, eigenvalue decomposition means for decomposing the covariance matrix created by the covariance matrix creation means to extract a plurality of eigenvalues and eigenvectors, and arriving at the array antenna based on the eigenvalues extracted by the eigenvalue decomposition means. Signal number detecting means for detecting the number of signals present, and a maneuvering vector a (θo-jΔθ) having the center of the arrival direction of the signal θo and the spread of the arrival direction Δθ as variables And a direction calculating means for calculating θo and Δθ corresponding to each arriving wave mode using an equation including the following equation and the eigenvector of the signal component of the received wave.

According to a fourth aspect of the present invention, the direction calculating means includes θo and Δθ
And θθ corresponding to the mode of each arriving wave are obtained by using an equation including the maneuvering vector a (θo-jΔθ) as an element and the eigenvector of the signal component of the received wave. That is, the maneuvering vector a (θo-j)
By using an algorithm obtained by modifying the conventional “ESPRIT” algorithm using (Δθ), the direction θo and the spread width 2Δθ of the arriving wave having spread can be obtained with high accuracy.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) One embodiment of a direction detecting method and a direction detecting apparatus according to the present invention will be described with reference to FIGS. This embodiment corresponds to claims 1 to 3. FIG. 1 is a block diagram showing the configuration of the direction detection device of this embodiment. FIG. 2 is a plan view showing an example of the positional relationship between the transmitting station and the receiving station. FIG. 3 is a graph showing a drawing example of a two-dimensional MUSIC spectrum.

In this embodiment, the sampling means, the covariance matrix creation means, the eigenvalue decomposition means, the signal number detection means and the direction calculation means of the second aspect are respectively a receiving circuit 21, a covariance matrix creation section 31, and an eigenvalue decomposition processing section. 32, wave number estimation processing unit 33 and two-dimensional MUSIC spectrum drawing unit 34
Corresponding to The spectrum drawing means of claim 3 corresponds to the two-dimensional MUSIC spectrum drawing unit 34.

In this embodiment, it is assumed that the direction of an incoming wave is detected in the situation shown in FIG. 2, for example. FIG.
Represents a case where the base station receives radio waves from one or a plurality of mobile stations. In a multipath environment such as an urban space, radio waves radiated from a mobile station are reflected or scattered by obstacles such as surrounding buildings. Therefore, the radio waves radiated from the mobile station reach the base station through different routes. Therefore, when the mobile station is viewed from the base station, the direction of the incoming wave spreads as shown in FIG. θo represents the angle of the center of the direction of the incoming wave, and 2Δθ represents the angle of the spread width in the direction of the incoming wave.

The direction detecting device shown in FIG. 1 can be used, for example, in the base station shown in FIG. 2 to detect the arrival direction θo and the spread width 2Δθ of radio waves arriving from each of a plurality of mobile stations. Note that the conventional "M
The USIC and ESPRIT algorithms do not assume that waves with a spread as shown in Fig. 2 will arrive, so it is necessary to accurately estimate the direction of the incoming wave even if these algorithms are used as they are. Can not.

In this embodiment, when a spread wave arrives using a method improved from the “MUSIC” algorithm, the arrival direction θo and spread width 2 of each wave are obtained.
Δθ can be detected with high accuracy. Referring to FIG. 1, the direction detecting device includes an array antenna 10, a receiving circuit 21, a signal processing circuit 30, a display device 41, and a printer 42. Array antenna 10 is composed of one of m that along the axial direction are arranged in a row at equal intervals at regular intervals d _x of the antenna element 11.

The same number of receiving circuits 21 as the number of antenna elements 11 are provided. Each of the receiving circuits 21 performs a receiving process of a signal received by the antenna element 11 connected to each of the inputs, and performs a predetermined time interval (T
(Seconds), sampling and A / D conversion are performed, and a digital signal having a complex amplitude is output. The output vector y (qT) of the digital signal output from the receiving circuit 21 is represented by the following equation.

(Equation 9) The output vector y (qT) is a value sequentially sampled at intervals of T seconds. The element number of the antenna element 11 corresponds to the element number of the output vector y (qT).
Further, where d number of incoming waves arriving from different directions, the spatial response of each _{v i (i = 1, ···} ,
It is assumed that the signal arrives with the complex amplitudes of d) and s _i (t).

The covariance matrix creation unit 31 extracts N snapshots from the output vector y (qT) obtained at the output of the receiving circuit 21, and creates a covariance matrix R thereof.
The covariance matrix R is represented by the following equation.

(Equation 10) The eigenvalue decomposition processing unit 32 receives the covariance matrix R output from the covariance matrix creation unit 31, and decomposes the covariance matrix R into a plurality of eigenvalues corresponding to the respective received signal components. In practice, m eigenvalues λ _i equal to the number of antenna elements 11 are obtained (i = 1,..., M).

Further, an eigenvector e _i corresponding to each eigenvalue is obtained (i = 1,..., M). The eigenvector is a weight vector used to extract each of the plurality of signal components from the received signal. Eigenvalue λ _i
It is assumed that the following relationship holds asymptotically.

[Equation 11] That is, the eigenvalue λ _i is divided into signal components (λ ₁ ,..., Λ _d ) and noise components (λ _{d + 1} ,..., Λ _m ) corresponding to each of the d arriving waves. Each signal component has a different magnitude from each other. All signal components are much larger than noise components.

Similarly, for the eigenvector e _i , d
Orthogonal bases Es (= [e
₁ ,..., _Ed ]) and the orthogonal basis En corresponding to the noise components
_{(= [E d + 1,} ···, e m]) is divided into a. These have the following features. The orthogonal basis Es asymptotically corresponding to the signal component matches a linear combination of a matrix V (= [v ₁ ,..., V _d ]) composed of d spatial responses as shown by the following equation.

(Equation 12) In addition, the spatial response v is calculated using the manipulation vector “a (θo−jΔ
θ) ”, En, which is the orthogonal complement space of the basis Es, is approximately orthogonal to the steering vector“ a (θo−jΔθ) ”. Further, the eigenvalue (λ _{d + 1} ,..., Λ _m ) corresponding to the noise component is asymptotically equal to the noise power because it is a projection onto the orthogonal space corresponding to the noise component of the covariance matrix R.

The eigenvalue decomposition processing unit 32 of FIG. 1, the obtained eigenvalues _{(λ 1, ···, λ m} ) gives the wave number estimation processing unit 33, the eigenvectors _{(e 1, ···, e m} ) a 2D MUS
This is given to the IC spectrum drawing section 34. The wave number estimation processing unit 33 checks the distribution state of the eigenvalues input from the eigenvalue decomposition processing unit 32 and estimates the number of incoming waves d. That is, the number of eigenvalues having a value much larger than the noise power is estimated as the number of arriving waves based on the relationship of the above equation (12). The wave number estimation processing unit 33 converts the estimated number of incoming waves d into a two-dimensional MU
This is given to the SIC spectrum drawing unit 34.

Two-dimensional MUSIC spectrum drawing section 34
Is based on the eigenvectors (e ₁ ,..., E _m ) input from the eigenvalue decomposition processing unit 32 and the number of incoming waves d input from the wave number estimation processing unit 33. The direction θo and its spread width 2Δθ are determined. As a method for obtaining the arrival direction of an incoming wave, a method using an orthogonal base corresponding to a noise component like the conventional "MUSIC" algorithm and a method using an orthogonal base corresponding to the signal component like the conventional "ESPRIT" algorithm are used. Method. Here, the central direction θo of the arrival direction of the arriving wave and its spread width 2Δθ are obtained by using a method improved from the “MUSIC” algorithm.

Two-dimensional MUS with θo and Δθ as variables
The function P (θo, Δθ) of the IC spectrum is expressed by the following equation.

(Equation 13) In equation (14), the two-dimensional MUSIC spectrum is obtained using the maneuvering vector “a (θo−jΔθ)” using θo and Δθ as variables and the orthogonal basis En corresponding to the noise component. An incoming wave can be estimated.

Actually, the two-dimensional MUSIC spectrum drawing section 34 draws the two-dimensional MUSIC spectrum of the above formula (14) as a visible graph on a plane in a form as shown in FIG. 3, for example. This graph is displayed on the display device 41.
Alternatively, it is output to the printer 42. The two-dimensional MUSIC spectrum drawing unit 34 can draw a graph as shown in FIG. 3 by calculating the value of the function P (θo, Δθ) while changing the variables θo and Δθ little by little. .

As shown in FIG. 3, a sharp peak appears in the two-dimensional MUSIC spectrum at a position corresponding to an actual arriving wave. From this peak position, the center θomin and the spread width Δθmin of the arrival direction of the incoming wave can be obtained as shown in FIG. The two-dimensional MUSIC spectrum drawing unit 34 actually draws a two-dimensional MUSIC spectrum including d peaks corresponding to each of d arrival waves. Therefore, the arrival direction and spread width of each arriving wave can be obtained from each peak position.

In this embodiment, the drawn two-dimensional MU
The case where the center θomin and the spread width Δθmin of the arrival direction of the arriving wave are obtained based on the SIC spectrum has been described. However, if a method such as searching for the root of a polynomial is used, the beak position of the function P (θo, Δθ) can be obtained. It is also possible to obtain it directly by calculation.

(Second Embodiment) Another embodiment of the direction detecting method and the direction detecting apparatus according to the present invention will be described.
This will be described with reference to FIG. This embodiment corresponds to claim 4. FIG. 4 is a block diagram showing the configuration of the direction detection device of this embodiment. This embodiment is a modification of the first embodiment. 4, elements corresponding to those in FIG. 1 are denoted by the same reference numerals.

The direction detecting device shown in FIG.
It has the same configuration as the apparatus of FIG. 1 except that an ESPRIT calculation processing unit 35 is provided instead of the IC spectrum drawing unit 34. For the same parts, the following description is omitted. The ESPRIT calculation processing unit 35 uses an equation expressed by the following equation (15) in order to directly obtain a mode corresponding to d arriving waves.

[Equation 14] Here, A in Expression (15) is a matrix composed of d manipulation vectors shown in Expression (13). It is. Also, J ₁ and J ₂
Is a (m−1) × m order matrix for selecting the first 1: (m−1) and second: m rows of the matrix A, respectively. Also, Φ
Is a matrix having the modes for d arriving waves in the diagonal terms. By obtaining the matrix Φ, the center θo and the spread width 2Δθ of the arrival direction of the incoming wave can be obtained.

In practice, the ESPRIT calculation processing unit 35 uses the aforementioned basis Es corresponding to the signal component instead of the matrix A. Based on the characteristics of the above equation (13), the following equation (17) holds asymptotically and approximately. Ψ in equation (17) is
It is a d-dimensional matrix represented by the equation. Due to the characteristics of the configuration of the matrix Ψ, the eigenvalues of Ψ match the modes of the d arriving waves.

(Equation 15) The ESPRIT calculation processing unit 35 performs the calculation processing as follows. First, the matrix Ψ is obtained by solving the equation (17). Next, the mode for the d arriving waves is calculated by obtaining the eigenvalue of the matrix Ψ. Finally, the center θo and the spread width 2Δθ of the arriving direction of the arriving wave are obtained using the above equations (19) and (20). Here, the symbols Re and Im represent a real part and an imaginary part, respectively.

In the conventional "ESPRIT" algorithm, since the direction of arrival is estimated using only the phase information of the mode corresponding to the incoming wave, it is not possible to estimate the incoming wave having a spatial spread. Was. ESPR
Since the IT calculation processing unit 35 performs the estimation using the amplitude information in addition to the phase information, it is possible to estimate the arrival direction and the spread width of the arriving wave having a spatial spread.

Although only the basic direction detecting device has been described in the first and second embodiments, the direction detecting method and the direction detecting device of the present invention are applied to various devices. be able to. For example, the present invention can be applied to a propagation characteristic measuring device, a mobile radio device, a wireless communication device, a radar device, a sonar device, a medical imaging device, a sound intensity measuring device, and the like.

[0036]

As described above, according to the direction detecting method and the direction detecting apparatus of the present invention, even if a wave having a spread in the direction of arrival arrives, the time series of the time series received by the array antenna can be obtained. It is possible to measure the center and the spread width of each incoming wave in the direction of arrival based on the signal.

Therefore, when the present invention is applied to, for example, mobile communication, the environment of each of a plurality of arriving waves can be accurately grasped at the base station, which is useful for improving communication quality.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a direction detection device according to a first embodiment.

FIG. 2 is a plan view showing an example of a positional relationship between a transmitting station and a receiving station.

FIG. 3 is a graph showing a drawing example of a two-dimensional MUSIC spectrum.

FIG. 4 is a block diagram illustrating a configuration of a direction detection device according to a second embodiment.

[Explanation of symbols]

 Reference Signs List 10 array antenna 11 antenna element 21 receiving circuit 30 signal processing circuit 31 covariance matrix creating unit 32 eigenvalue decomposition processing unit 33 wave number estimation processing unit 34 two-dimensional MUSIC spectrum drawing unit 35 ESPRIT calculation processing unit 41 display device 42 printer

Claims (4)

[Claims] 1. A direction detecting method for receiving one or a plurality of signals respectively arriving from arbitrary directions by an array antenna and detecting a direction in which each received signal arrives, comprising: The maneuvering vector a (θo−jΔ) having the center of arrival direction θo and the spread of the arrival direction Δθ as variables
θ), and calculating the function or equation based on the signals received by each element of the array antenna to obtain the center θo and the spread Δθ of the arrival direction of each signal. Direction detection method. 2. A direction detecting device for receiving one or a plurality of signals respectively arriving from arbitrary directions by an array antenna and detecting a direction in which each of the received signals arrives, wherein each of the array antennas Sampling means for sampling the received waves received by the elements at predetermined time intervals to obtain a predetermined number of signal values as a first matrix; and creating a covariance matrix for creating a covariance matrix from the first matrix Means, eigenvalue decomposition means for decomposing the covariance matrix created by the covariance matrix creation means and extracting a plurality of eigenvalues and eigenvectors, and arriving at the array antenna based on the eigenvalues extracted by the eigenvalue decomposition means Signal number detecting means for detecting the number of signals present, a steering vector a (θo-jΔθ) and a center θo of the arrival direction of the signal and a spread Δθ of the arrival direction as variables. And a direction calculating means for obtaining θo and Δθ necessary for maximizing the function using a function including the noise component En of the received wave as an element and the eigenvector of the noise component of the received wave. Characteristic direction detection device. 3. The direction detecting device according to claim 2, wherein a spectrum drawing means for displaying a change in the value of the function according to each change of the θo and Δθ in a two-dimensional space is provided as the direction calculation means. A direction detection device characterized by the above-mentioned. 4. A direction detecting device for receiving one or a plurality of signals arriving from arbitrary directions by an array antenna and detecting a direction in which each of the received signals arrives, wherein each of the array antennas Sampling means for sampling the received waves received by the elements at predetermined time intervals to obtain a predetermined number of signal values as a first matrix; and creating a covariance matrix for creating a covariance matrix from the first matrix Means, eigenvalue decomposition means for decomposing the covariance matrix created by the covariance matrix creation means and extracting a plurality of eigenvalues and eigenvectors, and arriving at the array antenna based on the eigenvalues extracted by the eigenvalue decomposition means Signal number detecting means for detecting the number of signals present, and a maneuvering vector a (θo-jΔθ) having the center of the arrival direction of the signal θo and the spread of the arrival direction Δθ as variables. A direction detecting device, comprising: a direction calculating unit that obtains θo and Δθ corresponding to each arriving wave mode using an equation included as an element and an eigenvector of the signal component of the received wave.