GB2142202A - Wide-base direction finder - Google Patents

Abstract

In a wide-base direction finder, equally spaced elements (1) of a circular antenna array are grouped in pairs. These element pairs are divided into at least two groups in such a way that all straight lines connecting the elements of the element pairs of a group are parallel to each other, and that the straight lines connecting the element pairs of one of the groups (I/VIII, II/VII, III/VI, IV/V) are normal to the straight lines connecting the element pairs of the other group (VI/VII V/VIII, I/IV, II/III). The differences in phase between the signals received by the elements of an element pair are measured. The measured values of the two groups are combined, and from the values so produced, the elevation-independent direction of arrival of a signal in the azimuth plane is determined. Evaluation of the all phase differences can take place simultaneously (as shown) or sequentially (Fig. 2). <IMAGE>

Description

SPECIFICATION Wide-base direction finder with circular antenna array The present invention relates to a wide-base direction finderwith a circular antenna array. A wide-base direction finder of this kind is disclosed in an article by G. Peuker, "ORTAC-M, ein neues TACAN-System", Elektrisches Nachrichtenwesen, Vol. 50, No. 4, 1975, pages 283-287.

In the facility described there, each radiating element is followed by a receiver branch. To obtain accurate measurement results, the amplitude and phase response of each receiver branch must be continuously monitored and, if necessary, varied. In addition, unambiguous results are obtained only if the spacing of the evenly distributed radiating elements does not exceed a given value. At this maximum spacing, however, mutual coupling between adjacent elements occurs. If element spacing is increased, the effects of mutual coupling are reduced. However, the measurement becomes ambiguous, so that additional measurements are required to resolve the ambiguity.

The invention seeks to provide a wide-base direction finder having a circular antenna arraywhich permits unambiguous angular measurements without additional ambiguity-resolving steps even if adjacent elements are widely separated.

According to the invention there is provided a wide-base direction finderwith a circular array of radiating elements, characterised in that the number of elements is an integral multiple of four, that the elements are grouped in pairs, that at least two groups are formed from the pairs of elements, such that all straight lines connecting the elements of the pairs of elements of one of the groups are parallel to each other, and that the straight lines connecting the pairs of elements of said one group are normal to the straight lines connecting the pairs of elements of the othergroupthatthe differences in phase between the signals received by the elements of each pair of elements are measured,thatthe measured values are fed in digital form to a computer, and thatthe computer determines the elevation-independent direction of arrival a of a signal in the azimuth plane liytheequation oc= arctan Y orbyan equation x corresponding to said equation,wherey is proportional to the direction of arrival in the azimuth plane determined with the first element group, and is proportional to the direction of arrival in the elevation plane determined with the second element group.

Thanks to the permissible large element spacing, the novel wide-base direction finder requires a relatively small number of elements. Mutual coupling between adjacent elements is weak. The novel wide-base direction finder is insensitive to interference caused by multipath propagation. Afinite extent ofthe radiation source and Doppler shifts of the frequency ofthe received signal have no effect on the measurement result. The same applies two phase and amplitude variations in the receiver branches. Depending on howfastthe bearing ofthe source isto be determined, one receiver pair is provided for each element pair or a single receiver pair is connected sequentially to different elements.

In orderthatthe invention and its various other preferred features may be understood more easily, two embodiments ofthe invention will now be described, byway of example only, with reference to the drawings, in which: Figure lisa schematic block diagram of a first embodiment of wide-base direction finder constructed in accordance with the invention, and Figure 2 is a schematic block diagram of a second embodiment of wide-base direction finder constructed in accordance with the invention.

Referring nowto Figure 1, a plurality of radiating elements 1 are equally spaced on a circle. In Figure 1, only eight elements 1, which are consecutively numbered with Roman numerals I to VIII, are shown for the sake of clarity. A possible real array, on which the arrays chosen in the following description are based, has a diameter of 57 A (A = wavelength of the signal whose direction of arrival isto be determined) and 32 elements. The smallest element spacing is thus 5.625 A, which ensures good isolation between adjacent elements.

The eightelements 1 of Figure 1 are divided into two groups, with each element belonging to both groups. In addition,the elements are grouped in pairs, such that the striaght lines connecting the elements of the pairs of a group are all parallel to one another. Thefirstgroup containsthe element pairs Viol, IIINIII, IVNII, and VNI, and the otherthe element pairsVIlNlIl, INI, IN, and III/IV,. The straight lines connecting the elements of the pairs of the second group are normal to the straight lines connecting the first group.Each elementisfollowed bytwo receivers 2, because each element belongs to both groups, as mentioned above.

The output signals of each receiver pair associated with an element pair are applied to a phase meter 3, which measures the difference in phase between the elements of the associated pair. The measured values are digitized in analogue-to-digital converters 4,then fed to a computer 5. The computer determines the elevation-independent angle of arrival in the azimuth plane in the manner described below. The angle of arrival is indicated in an indicator6,which is known perse.

Before dealing with the evaluation in detail, another embodiment will be explained with the aid of Figure 2. In this embodiment, only one receiver pair 28, 29 is present. This receiver pair is connected sequentiallytoall elementpairsofthetwogroupsvia switches 22,23,24,25,26,27. The receiver pair 28, 29 is followed by a phase meter 33 and an analogue-todigital converter30.Thedigitized measured values are fed to a computer 31. The latter not only determines the direction of arrival but also controls the switches.

In position a, the switches 26 and 27 are connected to the switches 22 and 23. The latter, in turn, connect the receiver pair 28,29 sequentiallyto the element pairs ofthe first group. In position b, the switches 26 and 27 are connected to the switches 24 and 25. The latter connectthe receiver pair 28,29 sequentially to the element pairs ofthe second group.

Thefollowing explains the evaluation in the computer. The evaluation is based on the following property: Each element pair represents an interferometer with a given spacing. The element pair is assigned a given radiation pattern orbeam.Theform ofthis beam depends on the spacing ofthe elements ofthe element pair. The signals received bythetwo elements are applied to two receivers. The beam points in the direction ofthe normal to the straight line connecting the two elements.This direction corresponds to the azimuth 0. If one of the two signals is applied to the associated receiver after being shifted in phase bythe value 0, the direction of the beam will be changed by a given angle, i.e. the beam will no longer point in the direction e = 0. This is utilized to scan the beam in steps through the azimuth range from 900 to +90 . Forthe embodiment, 512 steps are assumed. However, the phase shifts are not produced as real phase shifts but only taken into account in the computer. For the evaluation in the computer, y is set equal to sin 6. Thus, each y is assigned a given beam and, hence, a given angle of arrival 6.

In the representation of Figure 1, the first group of element pairs consists of eight elements and, thus, four element pairs. Consequently, the phase meters 3 measurefour phase differences (Pn (n = 1 to 4). The phase differences are #n = ssn - ssN-n+1 = 2Hn sin (&alpha;;) = 2Hng where N = number of elements ssn = phase of the signal received by the nth element relative to a reference phase Hn = spacing between the elements of an antenna pair in radians cc = the angle of arrival to be measured g =sin To determine the (elevation-dependent) angle of arrival in the azimuth plane, the product

orthe sum

is formed, with

whereat = amplitude of the signal received by the nth element = = direction of arrival in the elevation plane y = sine, w th 6 being a variable in the direction of arrival in the azimuth plane.

Fortheformation of the product, A, and qiarn variables which can be set equal to 1, as can be deduced. If N =32, one obtains N/2 = 16valuesfor un, from which the product U(y) is formed.

With thefirst element group alone, only conical coordinates would be obtained. It therefore suffices forthetime being to determine the direction of arrival ofthesignal in an azimuth range from -90 to +90 .

For an unambiguous determination, i.e. the determination ofthe direction ofarrival in the entire 360 range (direction of arrival in cartesian coordinates), the additional group is necessary.

With the computer, only a quantized measurement of the direction of arrival is possible. The measurement accuracy depends primarily on the diameter of the circular array and, duets the method ofevalua- tion, also on the numberof sample values. Forthe present example, in which 32 elements are used, it is advantageous to provide 512 sample values forthe +90 range. Each sample value corresponds to a given direction of arrival 6 and, thus, to a given beam, as indicated above. It is, therefore, necessary to form the product according to equation (1 ) 512times, the individual products differing from each other by different values of e (i.e. different beams).If the 512 calculated products are examined as a function of y, the largest product is obtained for that y = sin eat which e is equal to the angle of arrival cc. This y, which can assume positive or negative values, is used for the further evaluation.

The same calculation is then carried outforthe second group of element pairs. In this case, however, the product

is formed, where un(x) = Ancos (Hn (x-fcos #) un(x) = Ancos (Hnx - #n cos #) 2 wherex= cos e f=cos It is again determined at which xthe largest product is present. The angle of arrival cc is then cc = arc tans.

x The computational effort involved in calculating the direction of arrival in the manner described is great. However, there are ways of reducing this effort bystoring suitable values and using advantageous algorithms. As mentioned above, instead of the un(x) and u(y),the logarithm of these values can be computed. The computational effort is reduced by storing N/2 "logarithmic beam functions" in a memory, computing the shifts of the measured phases from the phase value for the direction of arrival 6=00, and adding the shifted "beam functions". By "beam functions" are understood the 512 beams for one element pairwhich coverthe azimuth rangefrom -90 to +90 .

To determine the direction of arrival oc in the computer, 2M evenly distributed points each having one beam assigned thereto are selected both in the rangeofxand in the range of y. The signal received via the nth beam can be described by U'n (k) = ln [cos (Hnk/M)] where k -M, +M.

lfthesource is not at azimuth 6 = 0 but at azimuth 6 = cm, the nth beam ofthefirst group has to be shifted by

and that ofthe second group hasto be shifted by

where M is 256forthe example described, and NINT isthe nearest integer.

Further, and

wherek,j = The (jAjn) and (kAkn) are computed mod (2M).

Forthe calculation of U'(j) and U'(k) it is not necessary to store the entire functions, onlythe progressive maxima and the values IMAX and kMAX corresponding to them.

The direction of arrivals in the azimuth plane can be computed from cos lilC exp(icc=) = (iMAX + ikMAX)1M where c is the direction of arrival in the elevation plane.

The accuracy of the angle of arrival computed by eitherofthetwo methods can be further improved by known methods of interpolation.

The angle of arrival can also be computed by other methods which will not be explained here, provided that suitable measured values are available as described.

For even greater accuracy, in addition to the original first and second groups of element pairs, a furtherfirstgroup and afurthersecond group are selected in such a waythatthe normal to the straight lines connecting the elements of an element pair of the first group points as precisely as possible (this depends on the number of element pairs) in the direction of arrival of the signal. This evaluation requires that the receiver pair be connected to the respective element pairs selected.

Accuracy can also be improved by combining the elements into several first and second groups and then averaging the measured values.

The evaluation requires thatthe assignment of the measured values to the element pairs be known in the computer. This will be the case if the connection of the desired element pairs to the receiver pair is controlled by the computer.

In the embodiment of Figure 1, it is possible to feed each measured phase value to one selected input of the computer, so thatthe assignmentofthe measured values to the element pairs is known in the computer.

Claims (7)

1. Awide-base direction finder with a circular array of radiating elements (1), characterised in that the number of elements (1) is an integral multiple of four,thatthe elements (1) are grouped in pairs, that at least two groups reformed from the pairs of elements, such that all straight lines connecting the elements ofthe pairs of elements of one ofthe groups are parallel to each other, and that the straight lines connecting the pairs of elements of said one group (INIII, IINII, IIINI, IVN) are normal to the straight lines connecting the pairs of elements ofthe other group (VINII, VNllI, lIlV, ll/lil), that the differences in phase (q)n) between the signals received by the elements of each pair of elements are measured, that the measured values are fed in digital form to a computer (5), and thatthe computer (5) determines the elevationindependent direction of arrival cc of a signal in the azimuth plane bythe equation oc = are tan Y or by an equation corresponding to said equation, where y is proportional to the direction of arrival in the azimuth plane determined with the first element group, and x is proportional to the direction of arrival in the elevation plane determined with the second element group.

2. Awide-base direction finder as claimed in Claim 1, characterised in that afterthe first measurement ofthe elevation-independent direction of arrival in the azimuth plane, a further measurement is made forwhich the element pairs are so chosen thatthe normal to the straight line connecting the elements of an element pair of a group points at least approximately in the direction of arrival.

3. Awide-base direction finder as claimed in Claim 1 or2, characterised in that a plurality of group pairs are formed sequentially, and that the measure mentresultsforthe individual grouppairsare averaged.

4. Awide-base direction finder as claimed in Claim 1 or 2, characterised in that each element pair is followed by a receiver pair (2).

5. Awide-base direction finder as claimed in Claim 1,2, or 3, characterised in that a single receiver pair is provided which is connected sequentially to the radiating elements.

6. A wide-base direction finder substantially as described with reference to Figure 1 of the drawings.

7. A wide-base direction finder substantially as described with reference to Figure 2 of the drawings.