JPH11248812A - Radio direction finder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a radio direction finder by which arrival directions of radio waves whose correlation is extremely high such as multipath waves can be separated over a wide frequency range. SOLUTION: Linear antenna elements 1, 2, 3 are arranged in the respective vertexes of an equilateral triangle so as to form a unit array 11. Similar unit arrays 12, 13 are arranged so as to constitute an equilateral triangle. Received waves by elements 1 to 6 are stored in a memory 23 as a digital series. After that, self-correlation matrixes R1, R2, R3 are computed respectively in the respective unit arrays 11, 12, 13 (24-1 to 24-3). Then, the correlation matrixes R1, R2, R3 are averaged in an average computing part 25. An evaluation function PMU (θ) by the MUSIC method is computed with reference to the averaged correlation matrixes (26). The angle θ of its peak value is searched in a search part 27, and the angle θ is displayed on a display part 28 as an arrival radio- wave direction.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a direction finder which uses an array antenna and applies the MUSIC method to detect directions of arriving waves from a plurality of directions even at the same frequency.

[0002]

2. Description of the Related Art Orientation measurement by the MUSIC method is disclosed in
No. 347529, Japanese Unexamined Patent Publication No. Hei 5-196716, "Direction finder" and the like. The direction detection by the MUSIC method cannot separate the arrival directions of radio waves having a large cross-correlation like multipath waves. From this point, it can be seen that the cross-correlation information for the multiple reflected waves can be suppressed by spatially averaging (moving average) the autocorrelation matrix in order to apply the MUSIC method for estimating the direction of arrival of spatially multiplexed waves. Engineering Research Association EMCJ89
−66, pp. 7-12, “Estimation of Direction of Arrival of Space Multiplex by Rotating Scanning of Antenna Using MUSIC Method”.

[0003] Direction detection by a linear array antenna
Is performed, the incoming
It is necessary to detect the direction of the radio wave. But the array
When the distance d between the elements is larger than the wavelength λ,
Wave direction θ₁, Θ_TwoSin θ in between ₁−sin θ_Two= Nλ / d
When the relationship (n is an integer) occurs, these radio waves are separated.
It is not possible. Therefore, there are many radio waves that cannot be separated.
Will be. That is, as shown in FIG.
Times, the interval D2 is set to 2 times as a unit (subframe).
Is repeated, the above-described problem of inseparability occurs.

[0004] From such a point, as shown in FIG.
After arranging the two sub-arrays of FIG. 3A, arranging the sub-arrays at an interval D2, that is, arranging the sub-arrays at unequal intervals, so that the phase relationships of all the sub-arrays do not coincide at the same time in 1997. Proceedings of the IEICE Communication Society Conference, 172 pages, B-2-34
"High resolution method DOA estimation using unequally spaced subarrays"
Has been proposed.

[0005]

However, even in the linear array antenna shown in FIG. 3B, it is impossible to separate the radio waves arriving at the same angle θ from both sides of the linear array antenna in the extension direction of the linear array as shown in FIG. 3C. . In addition, the unequally spaced sub-array method has a problem that the number of elements is increased as a whole.

[0006]

According to the present invention, at least three elements are two-dimensionally arranged to form a unit array, and at least three sets of such unit arrays are two-dimensionally arranged. The autocorrelation matrices of the received waves are calculated, the averages of the autocorrelation matrices are calculated, the MUSIC method is applied to the average autocorrelation matrix, the peak of the evaluation function is searched, and the direction of the peak is determined by the wave direction. It is estimated to be the direction of arrival.

[0007]

FIG. 1 shows an antenna portion of an embodiment of the present invention. In this embodiment, the linear antenna elements 1, 2, 3 are two-dimensionally arranged to form a unit array 11, and at least three sets of such unit arrays 11 are arranged two-dimensionally. In this embodiment, for example, a unit array 1 composed of linear vertical antenna elements 1, 2, 3
Numeral 1 is provided at three vertices (corners) of an equilateral triangle, and three elements 3, 4, and 5 are arranged at three vertices of an equilateral triangle of the same size to form a unit array 12.
The unit array 13 is formed by arranging 4 and 6 at three vertices of an equilateral triangle having the same size. That is, in the unit arrays 11, 12, and 13, each element is located at the vertex position of an equilateral triangle having the same element interval d, and the unit arrays 11 and 12 share the element 3, and the unit arrays 11 and 13 share the element 2 To
The element arrays 4 and 13 share the element 4 respectively. The elements 1, 5, and 6 are located at each vertex of a regular triangle having one side of 2d.

The unit arrays 11, 12, and 13 need not be connected to each other and may be separated from each other. In that case, the number of elements is increased by three in total. The unit arrays 11, 12, and 13 do not have to form an equilateral triangle, but preferably have the same shape. Further, the number of elements in one unit array is not limited to three, and may be four or more. However, the shapes of the elements must be the same.
In any case, it suffices if the reception waves of the corresponding elements can be added.

In FIG. 1, signals u ₁ (t), u
Assume that ₂ (t) comes from the directions of angles θ ₁ and θ ₂ .
These two waves are completely correlated and have the following relationship. _{u 2 (t) = αu 1} (t) ... (1) Now the coordinates of the elements _{1,2,3 (x i, y i)} (i =
1, 2, 3), the signal r1 _i (t) received by the elements 1, 2, 3 can be expressed by the following equation. _{r1 i (t) = Σ k} = 1 2 exp [-j (2π / λ) (x i sin θ k + y i cos θ k)] u k (t) + n i (t) ... (2) n i ( t) is the antenna noise voltage of the elements 1, 2 and 3.

Similarly, the signal r received by the elements 3, 4, 5
2 _i (t) can be expressed as follows using the coordinates of elements 1, 2, and 3. _{r2 i (t) = Σ k} = 1 2 exp [-j (2π / λ) {(x i + √ (3) R) sin θ k + y i cos θ k}] u k (t) + n i (t ) (3) R is the distance between the intersection point Wc of the line connecting the elements 1 and 4 and the line connecting the elements 2 and 5 and the element 3.

Similarly, the signal r received by the elements 2, 4, 6
3 _i (t) is represented by the following equation. _{r3 i (t) = Σ k} = 1 2 exp [-j (2π / λ) {(x i + √ (3) R / 2) sin θ k + (y i + 3R / 2) cos θ k}] u _k (t) + n _i (t) (4) Equations (2), (3), and (4) are each rewritten in a matrix expression.
When the relationship u ₂ (t) = αu ₁ (t) is used, the following is obtained.

R1 (t) = Au1 (t) + n (t) (5) r2 (t) = Au2 (t) + n (t) (6) r3 (t) = Au3 (t) + n (t) ... (7) u1 (t) = (u 1 (t) u 2 (t)) T ... (8) u2 (t) = [u 1 (t) exp [-j (2π / λ) √ (3) R sin θ ₁ ] u ₂ (t) exp [−j (2π / λ) √ (3) R sin θ ₂ ] ^T (9)

[0013]

(Equation 1)

A = (a (θ ₁ ) a (θ ₂ )) (11) When the correlation matrices of the equations (5), (6), and (7) are obtained, they are as follows.

[0015] ^{_{R 1 = AR u1 A H +}} σ 2 I ... (13) R 2 = AR u2 A H + σ 2 I ... (14) R 3 = AR u3 A H + σ 2 I ... (15) σ 2 is noise power , I are unit matrices. The conjugate transpose is represented by H and the complex conjugate is represented by *, and the signal correlation matrices _Ru1 , _Ru2 , R
_u3 is as follows.

[0016]

(Equation 2) E represents an expected value (time average).

[0017]

(Equation 3)

[0018]

(Equation 4)

Ν _i = exp [−j (2π / λ) √ (3) R sin θ _i ] (19) γ _i = exp [−j2π / λ (√ (3) R / 2 · sin θ _i + 3R / 2) Cos θ _i )] (20) The rank of the signal correlation determinants (16), (17), and (18) is clearly 1 and a correct orientation cannot be estimated even by applying the MUSIC method. Therefore, the arithmetic average of the equations (13) and (14), that is, the spatial average is calculated to suppress the correlation.

[0020]

(Equation 5)

[0021] At this time, even if the MUSIC method is applied to Expression (21), it is not possible to perform correct azimuth estimation. The condition under which the bearing cannot be estimated is ν ₂ −ν ₁ = 0 (22). Substituting equation (19) into equation (22) and rearranging it results in the following equation.

Sin θ ₂ −sin θ ₁ = nλ / (√ (3) R) (23) In the same manner as the integer, n is a condition under which the correct azimuth cannot be estimated by arithmetically averaging Expressions (13) and (15). Then, sin (θ ₂ + π / 3) −sin (θ ₁ + π / 3) = nλ / (√ (3) R) (24) Therefore, if there are no different θ ₁ and θ ₂ that simultaneously satisfy the expressions (23) and (24), the evaluation function is obtained by applying the MUSIC method, and the direction at which the peak value is obtained is always estimated as the correct direction. be able to.

Here, equations (23) and (24) are defined as simultaneous equations.
Then, by solving this, θ₁, Θ_TwoSought
You. sinθ₁= (-3A ± √ (9-3A)^Two)) / 6 (25) A = nλ / (√ (3) · R) By substituting equation (25) into equation (23), sin θ_Two
Is found. For example, at a frequency of 500 MHz and R = 0.5 m
About the case θ₁, Θ_TwoAnd calculate the uniqueness of each solution
The value (with the maximum eigenvalue being 1) was determined. In this case,
4 sets of θ₁, Θ _TwoWas found.

Number θ₁ θ_Two Specific value 1 6.4 ° 53.6 ° 1.0 1.34E-3 1.18E-3 2 173.6 ° 126.4 ° 1.0 1.67E-4 1.46E-4 3 233.6 ° 186.4 ° 1.0 1.27E-3 1.17E-3 4 306.4 ° 353.6 ° 1.0 1.68E-4 1.55E-4 where E-3 = 10^-3, E-4 = 10^-FourIt is. This result
In the result, θ of number 1 ₁And θ_TwoIs the number 3_Two, Θ
₁And 180 ° respectively, and the number 2 θ₁, Θ_Two
Is the number 4_Two, Θ₁And each is 180 ° different
You.

From the eigenvalues, the wave number is determined to be 1. According to a simulation experiment, three autocorrelation matrices R1, R2, and R3 are averaged (that is, a spatial means is performed), and the average value is calculated. By applying the MUSIC method, the evaluation function P
_{When the MU} was obtained, it was as shown by the solid line in FIG. As apparent from this figure, two peaks appear, and the orientations of these peaks are about 0.4 ° and about 55.6 °, which coincide with the true orientation. For other solutions, when obtaining the evaluation function P _MU, the true bearing was obtained in the same manner. Thus, it is understood that the azimuths of all incoming waves can be obtained according to the present invention.

The dotted line in FIG. 2 is obtained by averaging only R1 and R2 and obtaining the evaluation function _PMU for the corresponding one. In other words, this corresponds to the prior art. In this case, no peak appears on the _PMU line, and it can be seen that the direction of arrival of the radio wave cannot be measured. As understood from the above description, the reception signals of the antenna elements 1 to 6 are received by the reception units 21-1 to 21-6 as shown in FIG. 1B, respectively, and the AD converters 22-1 to 22- are respectively received. At 6, each is converted into a digital value sequence and temporarily stored in the reception memory 23. From these received digital value sequences, the autocorrelation matrix R for each of the unit arrays 11, 12, 13
1, R2 and R3 are represented by the formulas (13), (14) and (15)
Is calculated by the calculation units 24-1, 24-2, and 24-3.

Further, these autocorrelation matrices R1, R2,
The average (R1 + R2 + R3) / 3 of R3, that is, the spatial average, is calculated by the averaging unit 25. The MUSIC method is applied to the averaged autocorrelation matrix to calculate the azimuth evaluation function P _MU (θ) by the calculating unit 26. And calculates the azimuth θ of the peak in the evaluation function P _MU (θ) into the peak search unit 2.
7, and the direction of the searched peak is displayed on the display unit 28 as the direction of the incoming radio wave.

[0028]

As described above, according to the present invention,
Since three or more elements are used in a two-dimensional array (as a unit array), a phase difference occurs between these elements with respect to radio waves arriving from any direction, between at least two elements. As described with reference to FIG. 3, there is no problem that the linear array cannot be distinguished when it arrives at the same angle from both sides with respect to the arrangement direction.

Further, since at least three such two-dimensional unit arrays are two-dimensionally arranged, there are four sets of degeneracy of θ ₁ and θ _{2 according to} the principle described above. As described above, θ ₁ and θ ₂ satisfying sin θ ₁ −sin θ ₂ = nλ / d (n is an integer) are degenerated, and since there are countless numbers, according to the present invention, θ ₁ and θ ₂ are Only when it happens to be any one of the four sets, the distinction cannot be made. However, such a possibility is small, and in principle, most of the measurement can be performed correctly.

In addition, actually, if the evaluation function P _MU (θ) is obtained and its peak is searched for, it can be separated into two waves without the degeneracy of the four sets. Further, in the present invention, as shown in FIG. 1A, when elements are shared between unit arrays, when the value of the integer part of the number of arriving waves [n / 2] that can be separated the same (n is the number of antenna elements), for example, In the example of FIG. 1A, only six elements are required, and in contrast to the conventional eight elements shown in FIG. 3B, only two elements are required.

[Brief description of the drawings]

FIG. 1A is a diagram showing an antenna arrangement row in an embodiment of the present invention, and FIG. 1B is a block diagram showing a functional configuration of the embodiment of the present invention.

FIG. 2 MU when θ ₁ = 0.4 ° and θ ₂ = 53.6 °
The figure which shows the simulation result of the evaluation function value of a SIC method.

FIG. 3 is a diagram showing a linear array antenna of a direction finder to which a conventional MUSIC method is applied.

Claims (3)

[Claims] 1. A unit array comprising at least three elements two-dimensionally arranged to form a unit array, at least three sets of the unit arrays being two-dimensionally arranged, and means for calculating an autocorrelation matrix for a received wave for each of these unit arrays. Means for calculating the spatial average of the autocorrelation matrices of these unit arrays, and music (M
USIC): A direction finder including means for detecting the direction of arrival of a wave by applying the MultipleSignal Classification) method. 2. The direction finder according to claim 1, wherein at least one element of adjacent unit arrays is shared by both unit arrays. 3. The direction finder according to claim 2, wherein the unit array is provided with elements on each vertex of the equilateral triangle, and the three unit arrays are arranged so as to form an equilateral triangle. .