NL8203558A - LARGE BASE LEVEL. - Google Patents

Description

-1- * 4 *

Large base pillar.

The invention relates to a large base pillar with several antennas, which form a linear antenna with the length L, the antennas following each a receiver and a measuring device, which teliminate the phase of the received electromagnetic wave 5. measure relative to a reference phase, be connected, and in which the values measured by the measuring devices are processed in a processing device. Such a large base gauge is known from the German Offenlegungsschrift 27.17.997.

Large base gauges have the advantage of being relatively insensitive to interference caused by multipath expansion of received signals. In order to determine the incident direction of the electromagnetic wave, it must at least be known which phase has the received electromagnetic wave at the location of each antenna, based on a reference phase common to all antennas.

In addition to the phases, the amplitudes of the received electromagnetic wave are often also processed. It is further possible, instead of measuring the amplitude and the relative phase, to determine the I - ("In-phase portion") and Q - ("Quadrature portion") components of the received electromagnetic wave. A received signal is completely determined by the I-20 and Q components, respectively, by phase and amplitude values. Regardless of which information (phase; phase and amplitude; I and Q components) is to be processed, it is necessary to connect a receiver and a measuring device to each antenna.

25 In order to be able to realize unambiguous angles of attack via an azimuth range of + 90 °, it is necessary that the distance between adjacent antennas is less than half an operating wavelength (X).

Therefore, a large base gauge requires a large number of antennas, receivers and measuring devices. At these small antenna distances, the links between adjacent antennas further cause interference.

If the antenna distance is increased with the known large base gauges, then an expensive unambiguous method has to be carried out.

In addition, the likelihood that the particular directions of attack will be completely erroneous as a result of multiple road extensions increases.

It is therefore an object of the invention to indicate a large base pillar, 3203558% ί -2- in which an unambiguous bearing is possible even without an unambiguous method in a large angular range, when the distance between adjacent antennas is greater than half an operating wavelength.

With a large base pillar of the type according to the invention mentioned in the preamble, this is achieved in such a way that the antenna array contains m antennas, which are arranged as irregularly as possible within the antenna array, so that m is significantly smaller than n, where n = 2L / A + 1 (λ5 is the operating wavelength of the received electromagnetic wave) is that the processing device contains at least one calculator, which performs a fast Fourier transform of n complex input values composed of the m values measured in the measuring devices and (nm) constant values, the vectorial sum of which is equal to zero, and which are supplied to him in parallel or in series, that in the fast Fourier transform n output values are generated to which a specific space direction is added each time, which further is processed in the processing device it is determined which of the n output values of the calculator is the greatest in terms of value, and that the spatial orientation added to this output value is is the incident direction of the electromagnetic wave.

Advantageous further embodiments can be derived from the subclaims.

With the new large base gauge, it is possible without unambiguous method to unambiguously determine the angle of incidence of a received signal in a range of + 90 °, although the distance between adjacent antennas is greater than half an operating length of the received signal. If two of these large base gauges are provided, unambiguous angle measurements in the total azimuth range (360 °) are possible (conversion of conical to planar coordinates). The costs with regard to the antennas and the receiver and measuring devices connected to them are relatively small. The connections between adjacent antennas are minimal. The failure of a receiving branch consisting of antenna, receiver and measuring device is not critical.

The invention will be explained in more detail with reference to an example with reference to the drawings, in which: figure 1 shows a block diagram of the new large base pillar; Figure 2 shows an antenna for the large base pillar of Figure 1; and Figures 3 and 4 show envelope curves of the Fourier coefficients calculated at the new large base pillar.

40 The new large base pillar has an antenna row, the length of which is 8203558 * * 3 too L and which m has 11 antennae. Following each antenna, a receiver 12 and a measuring device 15 are connected. Mixing signals are sent to the receivers from a mixing oscillator 13 for converting the frequency of the received signals into the zero-intermediate frequency.

It is at least necessary for further processing or utilization that the phase of the received electromagnetic wave relative to a reference phase is known. Preferably, in addition to the phases, the amplitudes of the electromagnetic wave received by the antennas are also determined. In other ways, however, it is also possible to determine the I and Q components of the output signals of the receivers. The vector addition of the orthogonal I and Q components yields the vector describing the receiver output signal according to phase and amplitude.

For the further description it has been assumed that phases and amplitudes are measured. The measurements are made in the measuring devices 15, to which a signal is supplied for the phase measurement from a signal source 14, the phase of which forms the reference phase for all phase measurements. The measured phase and amplitude values are digitized in the analog / -20 digital converters 16 and supplied to a processor 19.

In the processing device, the incident direction of the electromagnetic wave is determined and fed to a display device (not shown).

The processing device 19 comprises a first calculator 17 for performing a fast Fourier Transform (FFT = Fast Fourier Transformation) and a second calculator 18, which determines which of the output values of the first calculator is the largest in value. Preferably, a special FFT calculator is used when very fast processing is to be done. When this fast processing is not necessary, these two calculators 17, 18 can then be replaced by a single calculator. This calculator then performs the fast Fourier transform and determines the largest output value.

As shown in Figure 1, it is assumed in the exemplary embodiment that a special calculator (17) is provided for the implementation of the fast Fourier transformation. It is known to the person skilled in the art that the data can be entered serially or in parallel in the calculator. For the purpose of faster understanding, a calculator is provided here, to which the data is entered in parallel.

40 The calculator which is described in the company brochure "SPS-1000 Super-Fast FFT Pipeline Processor" from Signal Processing Systems, Inc., 223 Crescent St., is suitable as a calculator for carrying out the fast Fourier transform 8203558 ί1 t 4. Waltham, MA 02154, USA.

N input values are applied to the FFT calculator 17 and the 5 calculator generates n output values.

For a better understanding of the processing principle it is first assumed (which is not the case with the new large base pillar) that the antenna line with the length L, n has antennas evenly distributed, n being chosen such that n ^ 2L / x + l (λ 10 is the operating wavelength of the received electromagnetic wave). The distance d between adjacent antennas is then equal to d £ λ / 2. With such a large base gauge, clear angle measurements are possible in an angle range of + 90 °.

The n complex input values Uyj of the calculator 17 contain in complex representation the phases and amplitudes measured in the measuring devices. For example, n fast values are now calculated in the fast Fourier transform.

These are: 20 n-1 F ° - Jo ** F, =. V uv exp (- ψ.,.

25 05 F = Vo exp (- iiSL) * v = 0 V n • '' ~ - 30 · ^ 1 F = Σ U exp (- l? 7 ™ (n-1)) n “1 v = 0 υ n 35 This system of equations can be summarized using the following relationship:

Fk = Vu exp (- AiHJS ,, 40 8203558 f% '5 where k and Ό can take values from 0 to n-1.

The expression given above is identical to the discrete Fourier transform (DFT) of the complex input values. The DFT can be performed in a particularly effective manner using FFT algorithms.

The n output values F ^ of the FFT calculator 17 correspond to the n signals, which will be obtained with an antenna system, the direction diagrams of which have n different (uniformly distributed) directions, that is to say that each Fourier coSfficient In mind, a radiation beam can be added in a specific space direction. Thus, a virtual radiation diagram is formed, which consists of n overlapping radiation beams. The fast Fourier transform is equivalent to a Butler radiation shaping matrix; see, for example, the "Radar Handbook" of N.I. Skolnik, Me Graw Hill, New York 15 1970, biz. 35-15 and the publications indicated there.

Since a specific spatial direction has been added to each of the n output values, it is only necessary to take into account which of the n output values is the largest in value for determining the incident direction. The unambiguous direction of attack ι / ’is then: 20 = arc sin for k ^ n / 2 resp. arc sin for * k> n / 2

Here are: 30 Ι / 'the incident direction of the received electromagnetic wave; λ the operating wavelength of the received electromagnetic wave; d the distance between adjacent antennas; n the number of the virtual radiation beams; k the number of the largest output value in terms of value.

35 The realization of such a large base gauge with n antennas required n receivers and n measuring devices, which is very expensive. In addition, due to the small distance of d * λ / 2 between adjacent antennas, disturbing antenna connections occur.

Subsequently, the new large-scale base-40 plumber is discussed again in the form to be realized.

8203558 6 • - ·

Surprisingly, it has been found that even then a range of + 90 ° overlapping virtual antenna bundles is obtained, whereby unambiguous soundings are possible if only m antennas are fitted instead of n antennas, where m is significantly smaller than n provided that the FFT calculator's dropped (nm) input values derived from measurements are replaced by zero (nm) values (or (nm) values whose vector sum is equal to zero). Thus, the FFT calculator further obtains n input values composed of m values, which are derived from measured values, and 10 (n-m) values, which are equal to zero.

Of course, the "dilution" of the antenna affects the Fourier coefficients at the output of the FFT calculator. Considering again the virtual many-beam diagram, the "dilution" of the antenna array causes an increase in the side peaks of the individual virtual beams of radiation. The main bundles, on the other hand, are only slightly changed.

With the "dilution" of the antennae, care must be taken to ensure that no separate imprints are created at the side peaks. The aim is to obtain antenna distributions, with the side peaks distributed as evenly as possible over the entire azinut range. It has been found that for this the distribution of the antennas within the antennae should be as irregular as possible.

Such a distribution is shown in Figure 2. If it is assumed that in order to obtain accurate polling results, an antenna line with a length of L 57X is provided, then without dilution and with d = 0.45 X, n = 128 receiving branches will now exist, consisting of antenna, receiver and measuring device.

The antenna is diluted such that the distances involved (in operating wavelength5) are 4.95; 4.05; 3.6; 3.15; 4.5; 4.05; 5.4; 4.95; eo 3.15; 3.6; 5.85; 5.4 and 4.5. With such antennas, the new large-base pillar, although the smallest distance is greater than 3.1 Λ, allows an unambiguous determination of the directions of incidence. Instead of n = 128 receiving channels, only m = 14 is needed and the coupling between adjacent antennas is limited.

The Fourier coefficients calculated for a given bearing direction can be derived from the diagram shown in Figure 3. For the sake of a better overview of the diagram, the measured values are not indicated per se, but the envelope curve of the measured values is indicated. It can be seen from the curve that at output 10 there is the greatest output value in terms of value 40, that is to say the signal to be sounded falls 7% from that direction in the manner associated with this Fourier coefficient. The incident direction is calculated according to the equation already indicated for the calculation of the incident direction l / * (with d = L / n-1).

With the aid of the further embodiment 5 described hereinafter, an improvement of the measuring accuracy and a reduction of the sensitivity against disturbances due to multiple-way extensions is obtained.

Further m input values are derived from the first input values obtained from the measured values. These further input values differ from the first input values with a certain phase difference.

The (n-m) values that are equal to zero remain unchanged. With these new values a further fast Fourier transformation is carried out, ie one obtains n further Fourier coefficients and with that also n further virtual radiation beams. If these further virtual radiation beams must always lie between the first virtual radiation beams, the phase shift of the individual input values must be equal to T. V / n (Ό is the position of the relevant antenna element in the thought grid of n -20 equal positions; V equals 0.1 ... n-1).

In this case, instead of n = 128 Fourier coefficients, 2n = 256 Fourier coefficients, i.e. also 256 virtual radiation beams, are now obtained. The direction of incidence is determined again by determining the largest Fourier coeff value in terms of value. The envelope of the calculated Fourier coefficients is shown in figure 4.

An increase in the measuring accuracy is also possible with the further embodiment described below, according to which an interpolation occurs.

Using the values of the Fourier coefficients Fj ^ j 30 and Fjj -_ ^, where F ^ is the largest Fourier coSefficient in value, a correction value Δ = C In (A / B) (with c = constant; A, B = amplitude values of the Fourier coefficients F ^ -j. ^, F ^ -i). The precise direction of attack xT is then: 35 ^ = arc sin HslAlA for (k + Δ)> n / 2 resp. nd if '= arc sin for (k + Δ) <n / 2 40 3203558 4 ^ 8 (For the antenna in question, c = 0.431).

The current description assumes that a single linear antenna array has been installed. This means that only an unambiguous angle range of up to + 90 ° can be monitored. In addition, due to the symmetry of the row, the angle information is obtained in a conical coordinate system. Different directions of incidence cannot be distinguished here, when they lie on a conical mantle which is rotationally symmetrical relative to the antenna array.

Unambiguous measurement results over the total azimuth range of 360 ° 10 are obtained when a second antenna is provided, which is arranged perpendicular to the first antenna. With the aid of both antennae angles of incidence and in conical coordinates can be measured. From these values, the planar azimuth incident direction y5 can be calculated according to and, if desired, the elevation incident direction according to 20 Θ = arc cos \ f (sin sin ϊ ^ 2) 9 * * 8203558

Claims (6)

1. Large base gauge with a plurality of antennas, which form a linear antenna of length L, the antennas each following a receiver and a measuring device, which have at least the phase of the electromagnetic wave received with respect to a reference phase are connected, and the values measured by the measuring devices are processed in a processing device, characterized in that the antenna array contains antennas (11) which are arranged as irregularly as possible within the antenna array (Figure 2), that m is significantly less than n, where n - 2LA + 1 (λ is the operating wavelength of the received electromagnetic wave), that the processing device (19) contains at least one calculator (17) that provides a fast Fourier transform (FFT) of n outputs complex input values composed of the m values measured in the measuring devices (15) and (nm) constant values, the vectorial sum of which is equal to zero, and which is parallel or in series aa n are supplied to him that in the fast Fouier transform n output values are generated to which a specific spatial direction is added each time, which further determination is made in the processing device (18, 19) which of the n output values of the calculator (17) in terms of value is the greatest, and that the spatial direction added to this output value is the incident direction of the electromagnetic wave. A large base gauge according to claim 1, characterized in that the (n-m) constant signals are equal to zero. Large base gauge according to claim 1 or 2, characterized in that further complex values are derived from the first m complex input values m supplied to the calculator (17), which differ from the first m values with a constant phase shift, which with these further m complex values a further fast Fourier transformation is carried out, wherein n output values are again generated to which a specific spatial direction is added each time, which in the processing device is determined which of the 2n output values which were generated in the fast Fourier transformation, is the greatest, and that the spatial direction added to this output value is the incident direction of the electromagnetic wave. Large base gauge according to claim 1, 2 or 3, characterized in that, in order to increase the measuring accuracy of the incident direction from the output values, which are adjacent to the largest value in terms of value, a correction value is determined, which is calculated in the calculation of the 40 approach direction is taken into account. 8203558 A large base gauge according to any one of claims 1 to 4, characterized in that the complex input values supplied to the fast Fourier transform executing calculator (17) contain, in addition to the measured value values, the corresponding amplitude values. The large base gauge according to any one of claims 1 to 4, characterized in that the I-components ("in-phase portion") and Q determined in the measuring devices are determined by the calculator performing the fast Fourier transform (17). components ("quadrature portion") are supplied, and that the Fourier transform is performed with this complex value. 0 ********** 8203558