METHOD AND DEVICE FOR GENERATING A MODULATED CONTINUOUS-WAVE RADAR SIGNAL

Abstract

A method for measuring an elevation angle and/or azimuth angle with an antenna array. Identical transmitted signals that are formed of successive linear-frequency-modulated ramps are transmitted through the transmitting antennas of the antenna array using time division multiplexing, wherein the time division multiplexing is achieved through alternating attenuation of the signals transmitted by the transmitting antennas. Echoes of the transmitted signals are received by the receiving antennas and are down-converted to a baseband and sampled. The down-converted and sampled echoes are transformed by an FFT into a 2D image domain. Phase differences are determined from the image data, and, in order to compensate for a systematic error present because of the lack of separation of the two transmitted signals, an error-compensated elevation angle and/or an error-compensated azimuth angle is determined by means of a compensation.

Claims

1. A method for measuring an elevation angle and/or an azimuth angle with an antenna array, the antenna array comprising at least two transmitting antennas that have a horizontal and a vertical spacing from one another, at least four receiving antennas that have a horizontal spacing from one another, and a monolithic microwave circuit, the method comprising: transmitting identical transmitted signals formed of successive linear-frequency-modulated ramps through the transmitting antennas of the antenna array using time division multiplexing, the time division multiplexing being achieved through alternating attenuation of the signals transmitted by the transmitting antennas; receiving echoes of the transmitted signals by the receiving antennas; and down-converting the received echoes to a baseband and sampling the received echoes; transforming the down-converted and sampled echoes via an FFT into a 2D image domain; determining phase differences from the image data; and determining an error-compensated elevation angle and/or an error-compensated azimuth angle via a compensation in order to compensate for a systematic error present because of the lack of separation of the two transmitted signals, the compensation being performed: after the measurement of the phase differences, or after the measurement of the phase differences and the calculation from the measured phase differences of an azimuth angle or elevation angle that is erroneous because of the lack of separation of the transmitted signals.

2. The method according to claim 1, wherein the compensation uses a priori knowledge about the systematic error or its effects.

3. The method according to claim 2, wherein the a priori knowledge is implemented in the form of values stored in a memory.

4. The method according to claim 3, wherein values are entered in a lookup table, from which they are read for the compensation.

5. The method according to claim 2, wherein the a priori knowledge is implemented in the form of an equation or multiple equations, which are used to calculate a compensated elevation angle and/or azimuth angle.

6. The method according to claim 5, wherein the a priori knowledge is contained in coefficients of the equation or equations, among other things. The method according to claim 5, wherein the phase differences are variables of one or more equations.

8. The method according to claim 3, wherein the phase differences are input quantities for lookup of compensated values in a lookup table.

9. The method according to claim 2, wherein the phase differences are initially used for calculating an erroneous elevation angle and/or an erroneous azimuth angle.

10. The method according to claim 9, wherein the erroneous angle or angles are variables of equations for calculating one or more compensated values and/or input quantities for lookup of compensated values in a lookup table.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The present invention will become more fully understood from the detailed description given hereinbelow and the accompanying drawings which are given by way of illustration only, and thus, are not limitive of the present invention, and wherein:

[0038] FIG. 1 shows a conventional linear frequency-modulated continuous wave radar signal;

[0039] FIG. 2 shows a conventional antenna array in a schematic representation;

[0040] FIGS. 3a and 3b show systematic errors of the elevation angle and of the azimuth angle without compensation;

[0041] FIG. 4 shows a dependence of the actual elevation angle on the erroneous elevation angle {circumflex over ()}.sub.E and the erroneous azimuth angle {circumflex over ()}.sub.A;

[0042] FIG. 5 shows the average correction error for various step sizes in the second method according to the invention;

[0043] FIGS. 6a and 6b show the dependence of the accuracy of the approximation from the third method according to the invention on the quantization of the coefficients and the order of the polynomial; and

[0044] FIGS. 7a to 7c show the systematic estimation error in compensation from the first method according to the invention.

DETAILED DESCRIPTION

[0045] According to an exemplary embodiment, the errors of the azimuth angle and/or of the elevation angle are reduced iteratively.

[0046] In a first step of the iterative method, first erroneous uncompensated azimuth angles .sup.0.sub.A and elevation angles .sup.0.sub.E are determined from the measurements .sub.1, 66 .sub.2, the values d.sub.Tx, d.sub.R, d.sub.Ty determined by the antenna array, and the wave number k of the electromagnetic wave, with the aid of the above-mentioned Equations (1) and (2).

[0047] With the aid of a priori knowledge, a first compensation value K(.sup.0.sub.A, .sup.0.sub.E) is then determined. The determination of the first compensation value is accomplished by means of a calculation or by readout from a memory in which a lookup table can be stored.

[0048] Using the compensation value K(.sup.0.sub.A, .sup.0.sub.E), a first compensated elevation angle .sup.1.sub.E is then calculated with the following equation:

.sup.i.sub.E=.sup.0.sub.E+K(.sup.i1.sub.A, .sup.i1.sub.E) (4)

[0049] The associated compensated azimuth angle .sup.i.sub.A is then determined from the equation derived from Equation (2):

[00003]$\begin{array}{cc}{}_A^i=\frac{{}_1}{\cos \left({}_E^i\right)}.fwdarw.{}_A^i={\sin }^{-1}\left(\frac{{}_A^i}{k.Math.d_R}\right)& \left(5\right)\end{array}$

[0050] The compensated elevation angle calculated with the first compensation value and the first compensated azimuth angle calculated therewith are still erroneous. They can be used as input quantities for a second compensation step in which a second compensation value is determined through a second calculation or a second readout from a memory.

[0051] Using the second compensation value, the second compensated elevation angle can then be calculated using Equation (4). The second compensated azimuth angle then results from use of Equation (5). Further compensation steps can follow.

[0052] The method can be continued iteratively until the error is minimized such that further processing of the compensated azimuth angles and elevation angles is then reasonably possible.

[0053] It has been demonstrated in an investigation that even two iterations are sufficient to largely compensate the systematic estimation error resulting from the coupling of the transmitting antennas. This is shown in FIGS. 7a to 7c.

[0054] According to another exemplary embodiment, an erroneous elevation angle and an erroneous azimuth angle are first calculated from the measured phase differences by means of Equations (1) and (2). These erroneous quantities are used as input quantities for reading a compensated elevation angle out of a memory in which a lookup table can be stored. A compensated azimuth angle is then determined by means of the compensated elevation angle and the of Equation (2).

[0055] The lookup table from which the compensated elevation angle can be read is based on measurements. For this purpose, a space around the antenna array can be sampled as finely as possible with the aid of a strong reflector, for example, which means that the reflector is displaced between two samples by an angular amount in height (elevation) or in the plane (azimuth). From the phase differences .sub.1, .sub.2 measured in this process, erroneous elevation angles {circumflex over ()}.sub.E and azimuth angles {circumflex over ()}.sub.A based on the measurements are calculated by means of Equations (1) and (2). The actual elevation angle of the measurement arrangement can be uniquely assigned to these erroneous angles. This assignment of the erroneous elevation angle {circumflex over ()}.sub.E and the erroneous elevation azimuth angle {circumflex over ()}.sub.A to the actual elevation angle is then stored in the lookup table.

[0056] The dependence of the actual elevation angle on the erroneous elevation angle {circumflex over ()}.sub.E and the erroneous elevation azimuth angle {circumflex over ()}.sub.A is shown graphically in FIG. 4.

[0057] A A

[0058] If erroneous elevation angles {circumflex over ()}.sub.E and azimuth angles {circumflex over ()}.sub.A are produced later based on the measurement of the phase differences .sub.1, .sub.2, the actual elevation angle can be read out of the lookup table stored in a memory. The actual azimuth angle can then be determined by means of Equation (2).

[0059] The measurements for determining the lookup table can be carried out for each antenna array, for example during a so-called EOL calibration. The measurement is then independent of production and component tolerances. However, great calibration effort is then required.

[0060] Alternatively, measurements could also be performed on a sample of antenna arrays, the results of which are then applied to all antenna arrays.

[0061] A lookup table is shown graphically in FIG. 4.

[0062] The single-valued region in the elevation direction has been chosen as +/30, which is to say 60. It is evident in FIG. 4 that a symmetry is present in the opposite quadrants. This can be used to reduce the memory requirement. Since the table thus generated contains only discrete values, the determination of correction values at intermediate points takes place through interpolation, for example through linear interpolation.

[0063] Despite the utilization of symmetry, a large memory is nonetheless necessary in order to store the lookup table with adequate accuracy.

[0064] For example, if one assumes an angular resolution (step size) of 1 in the azimuth and elevation angle directions, and a 16-bit quantization of the values, i.e., 2 bytes per value, then with an angular range of +/30 for elevation and +/90 for the azimuth angle, the result is a memory requirement of 180*60/2*2 bytes=10.8 Kbytes. If one reduces the resolution from 1 to 5, the memory requirement can be reduced to 432 bytes at the expense of accuracy. The average correction factor for various step sizes is shown in FIG. 5.

[0065] In another exemplary embodiment, the dependence of the actual elevation angle on the erroneous elevation angle {circumflex over ()}.sub.E and the erroneous azimuth angle {circumflex over ()}.sub.A, as is graphically represented in FIG. 4, is approximated by a polynomial. The coefficients that are obtained through this approximation can be stored in a memory. This will generally take place once during setup of the antenna array. The coefficients and the polynomial constitute the a priori knowledge that is used in the third method for compensation of the error.

[0066] It has been shown that the accuracy of the determination of the coefficients, in particular the quantization and the order of the polynomial selected for the approximation, has a great effect on the quality of the approximation (see FIG. 6). In order to achieve a sufficiently accurate approximation, the inventor proposes a 5.sup.th order polynomial for the two variables of the polynomial (i.e., in both angular directions), and a 32-bit quantization of the coefficients. This means that 21 coefficients of 32 bits are stored. This results in a memory requirement of approximately 48 bytes (21*4 bytes). The memory requirement is reduced by approximately 90% as compared to the second method with 5 angular resolution. However, this comes at the expense of an increased computing load for every raw target, since in this case a 5.sup.th order polynomial must be evaluated for every raw target, resulting in 70 multiplications and 20 additions.

[0067] The invention being thus described, it will be obvious that the same may be varied in many ways. Such variations are not to be regarded as a departure from the spirit and scope of the invention, and all such modifications as would be obvious to one skilled in the art are to be included within the scope of the following claims.