Radar measuring method

Abstract

To provide an improved method for radar measuring.SOLUTION: A method and a device for determining a value of relative speed of a radar target include FMCW radar measurement in which a transmission signal subjected to frequency modulation in ramp form is transmitted. A modulation pattern of the transmission signal includes a first sequence of ramps performed in series by being temporally shifted at a certain time interval, and at least one additional sequence of ramps performed in series by being temporally shifted at the same time interval. The sequences are temporally entangled with each other. Further, in the method and the device, a received signal is downconverted to a baseband signal, and a two-dimensional spectrum is calculated according to two-dimensional Fourier conversion separately with respect to the sequences respectively, from the baseband signal. At this point, conversion is performed for each ramp in a first dimension, whereas conversion is performed for a ramp index in a second dimension, in order to determine a value of relative speed of a radar target on the basis of at least two peak positions in the spectrum of the two-dimensional spectrum of the baseband signal.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention provides a method for determining the relative velocity of a radar target, comprising: (a) an FMCW radar measurement is performed, in which a ramp-like frequency modulated transmit signal is transmitted, the modulation pattern of the transmit signal including a first sequence of ramps spaced apart in time by a time interval and at least one further sequence of ramps spaced apart in time by the same time interval, the sequences interdigitated in time; (b) downconverting the received signal to a baseband signal, and calculating a two-dimensional spectrum from the baseband signal separately for each of the sequences by a two-dimensional Fourier transform, transforming by ramp in a first dimension and by ramp index in a second dimension; and (c) determining a value of the relative velocity of the radar target based on the location of peaks in at least two spectra of the two-dimensional spectrum of the baseband signal.

[0002] The invention further relates to a radar sensor, in particular for motor vehicles, which is designed to carry out the method. [Background technology]

[0003] In automobiles, FMCW radar sensors are used to capture the traffic environment, particularly to determine the position of other vehicles, which can then be used for various assistance functions, such as automatic distance control, automatic collision warning, or even to automatically initiate an emergency braking maneuver if a collision is imminent.

[0004] In FMCW (Frequency Modulated Continuous Wave) radar sensors, a transmit signal is used whose transmission frequency is modulated in a ramp shape and is transmitted continuously during the ramp. The received signal is mixed with the transmitted signal to generate a baseband signal, which is then sampled and evaluated.

[0005] The frequency of a baseband signal corresponds to the frequency difference between a signal transmitted at a given time and a signal received at the same time. Due to frequency modulation of the transmitted signal, this frequency difference depends on the time of flight of the signal from the radar sensor to the object and back, and therefore on the distance between the objects. However, due to the Doppler effect, the frequency difference also includes a portion caused by the relative velocity of the objects. Therefore, measuring the frequency difference on a single ramp does not yet allow for the determination of distance and relative velocity, but only provides a linear relationship between these quantities. This relationship can be shown as a straight line in a distance-velocity graph (dv graph).

[0006] FMCW radar sensors are known that use sequences of identical, relatively short ramps, so-called "rapid chirps," which have a large frequency shift compared to their duration and are therefore so steep that the interval-dependent portion of the frequency shift dominates in the baseband signal. In this case, the Doppler shift is determined by sampling the ramp sequence.

[0007] To enable an unambiguous determination of the relative velocity within the desired measurement range of the relative velocity, a sufficiently high repetition rate of the ramps is required, in particular the time delay between successive short ramps must be less than half the period of the Doppler frequency.

[0008] In order to enable accurate velocity and separation estimation of radar objects with as low hardware and computational costs as possible, it is proposed to use multiple sequences of frequency-modulated ramps with the time interval between the ramps being such that the ramp sequences undersample the Doppler shift, resulting in ambiguous information about the relative velocity.

[0009] From DE 10 2014 212 280 A1 a method of the type mentioned at the beginning is known, in which determining possible relative velocity values ​​of the radar target based on the locations of peaks in the at least one two-dimensional spectrum of the baseband signal, the values ​​being periodic with a predetermined velocity period; checking the phase relationships of spectral values ​​obtained at the same locations in the separately calculated two-dimensional spectra for agreement with predicted phase relationships for a plurality of periodic values ​​of the determined relative velocity; and From the determined periodic values ​​of the relative velocity, an estimate of the relative velocity of the radar target is selected based on the result of the check, thereby resolving the ambiguity. [Prior art documents] [Patent documents]

[0010] [Patent Document 1] DE102014212280A1 [Patent Document 2] DE102014122284A1 [Patent Document 3] DE102017200317A1 [Patent Document 4] DE102014223990A1 [Non-patent literature]

[0011] [Non-Patent Document 1] HLVan Trees, "Optimum array processing - Part IV of detection, estimation, and modulation theory," John Wiley, 2002 [Non-patent document 2] J. Capon, "High-Resolution Frequency-Wavenumber Spectrum Analysis," Proceedings of the IEEE, 1969, Vol. 57, pp. 1408-1418. [Non-patent document 3] H. Cox, R. Zeskind, and M. Owen, "Robust adaptive beamforming," IEEE Transactions on Acoustics, Speech and Signal Processing, vol. 35, no. 10, pp. 1365-1376, 1987. Summary of the Invention [Problem to be solved by the invention]

[0012] The object of the present invention is to provide a method of the type mentioned at the outset which allows for better detection results to be achieved. [Means for solving the problem]

[0013] According to the present invention, this problem is solved in that each of the two-dimensional spectra formed for the different sequences is modified by multiplication with a beamforming function that depends on the velocity variable and the time offset of the corresponding sequence relative to the first of these sequences, and the relative velocity of the radar target is determined based on the coherent sum of the modified spectra.

[0014] The beamforming function allows targeted variation in the radar sensor's sensitivity to objects with specific relative velocities. For example, the radar sensor can be made more sensitive to a vehicle ahead, thereby increasing the radar sensor's range and signal-to-noise ratio for this particular object class. Because a vehicle ahead will typically have a relative velocity near zero, the beamforming function is selected so that the radar sensor is most sensitive to objects with a relative velocity of zero.

[0015] The functional principle is similar to that of beamforming when transmitting and receiving radar signals using an antenna array with multiple offset antenna elements. When transmitting radar signals, the antenna elements are supplied with transmit signals, the phases of which are shifted by a phase difference proportional to the offset of the antenna elements. As a result, the maximum transmit power is not emitted in the direction normal to the array of antenna elements, but in a direction that forms an angle with this normal that depends on the phase offset. When receiving radar signals, the direction of maximum sensitivity of the radar sensor can also be determined by multiplying the signals received by each antenna element by a phase factor proportional to the offset of the antenna elements. In this case, the beamforming function is a function of the angular variable that determines the direction of maximum sensitivity and the spatial offset of the antenna elements.

[0016] The proposed method increases the sensitivity of the radar sensor to a specific relative velocity based on the principle of using a similar beamforming function, which accordingly depends on the velocity variable and the time offset of the sequence of frequency ramps.

[0017] The beamforming function may also be selectively configured to desensitize the radar sensor to objects with certain relative velocities. This may be utilized, for example, to simplify and improve object separation in multi-target scenarios by desensitizing the radar sensor to stationary objects that are not important for spacing control, i.e., objects whose relative velocity is the inverse of the absolute velocity of the ego vehicle, and / or to facilitate disambiguation in determining relative velocity.

[0018] In conventional methods, resolving ambiguity in measuring relative velocity is particularly difficult when undersampling in the Doppler dimension results in overlapping radar targets with different relative velocities. In this case, to resolve the ambiguity, a large number of ambiguity hypothesis combinations under consideration must be checked in the two-dimensional spectrum, which requires very high computational costs. In the method according to the present invention, the number of hypotheses to be checked can be significantly reduced by reducing the sensitivity to stationary objects. This allows for a greater degree of undersampling and a corresponding increase in the ramp time interval, thereby reducing computational costs and the required computing power of the hardware.

[0019] The method according to the invention also has the advantage that its relatively low sensitivity to stationary targets reduces the magnitude of peaks attributable to these targets in the spectrum, thereby simultaneously reducing the frequency and extent of peak overlap, which may, among other things, reduce the frequency of situations in which a peak in a particular cell in dV space is obscured by a very prominent peak in an adjacent cell.

[0020] Advantageous embodiments of the invention are set out in the dependent claims. Preferably, within each sequence, successive lamps have the same lamp slope and the same lamp center frequency difference, particularly preferably the same frequency shift, optionally with a lamp center frequency difference not equal to zero, and lamps with the same lamp index within each sequence have the same lamp slope and the same lamp center frequency, particularly preferably the same frequency shift. If the frequency progression of all lamps in all sequences is identical, up to a selected lamp-to-lamp frequency difference optionally not equal to zero, the phase relationship resulting from the relative velocity of the radar target can be measured particularly accurately.

[0021] If the time offset between sequences and the time interval between ramps within a sequence are of the same magnitude, the available measurement time can be utilized particularly well. Furthermore, the influence of object acceleration on the phase relationship between the baseband signals of the individual sequences can be kept as small as possible. Furthermore, suitable values ​​can be selected for the time offset between sequences and the time interval between ramps within a sequence that are as "incommensurable" as possible, i.e., not multiples of each other. In this case, the resolution of ambiguity results in a particularly large measurement range for the relative velocity. Accordingly, the modulation pattern includes pauses between ramps. In particular, it is preferable for the modulation pattern to have at least one pause that is regularly repeated between each two successive ramps of a sequence, with the time interval from pause to pause being the same as the time interval between ramps of a sequence.

[0022] Preferably, the lamps of each sequence are alternated over the main period of the modulation pattern, i.e., the sequences overlap to a large extent in time. The time offset between the lamps of each further sequence and the corresponding lamps of the first sequence is preferably less than twice the time interval between the lamps in the respective sequence, and particularly preferably less than this time interval. The latter means that one lamp of each further lamp sequence is always transmitted between two successive lamps of the first sequence.

[0023] Further examples of modulation methods and modulation patterns that can be combined with the proposed method are described in DE 102014122284 A1 and DE 102017200317 A1.

[0024] Optionally, angle estimation may be added to the determination of object spacing and relative speed. An example of a suitable method is described in DE102014223990A1. The proposed method is particularly suitable for radar sensors in which radar measurements are performed in time division multiplexing or Doppler division multiplexing with different combinations of transmit and receive antennas. In this case, velocity beamforming and coherent spectral addition are performed for each transmit antenna. The spectra can then be added noncoherently or coherently (by spatial beamforming) for the transmit and receive antennas. In the case of Doppler division multiplexing, transmitter assignment can also be performed following velocity disambiguation, for example, based on the method described in DE 10 2017 2003 17 A1.

[0025] In the following, exemplary embodiments are explained in more detail on the basis of the drawings. [Brief explanation of the drawings]

[0026] [Figure 1] FIG. 1 is a block diagram of an FMCW radar system. [Figure 2] FIG. 10 shows a modulation pattern in which two sequences of identical lamps are transmitted with a time offset T12. [Figure 3] 1 is a velocity / interval graph in which the relative velocity of a radar target takes on periodic values ​​at predetermined intervals. [Figure 4] 10 is a graph of a likelihood function for a parameter of relative velocity. [Figure 5] FIG. 2 is a more detailed block diagram of the evaluation mechanism of the FMCW radar sensor. [Figure 6] FIG. 1 is a flow diagram of a method according to the present invention. [Figure 7] 4 is a graph showing the dependence of the sensitivity of a radar sensor for different speeds. [Figure 8] FIG. 8 is an enlarged detail of FIG. 7. [Figure 9] FIG. 8 is an enlarged detail of FIG. 7. DETAILED DESCRIPTION OF THE INVENTION

[0027] FIG. 1 shows a simplified block diagram of an FMCW radar sensor 10, which is mounted, for example, in the front of a vehicle, and serves to measure the distance d and relative velocity v of objects 12, 14, e.g., vehicles traveling ahead. The radar sensor 10 includes a voltage-controlled oscillator 16 that provides a frequency-modulated transmit signal via a mixer 18 to a transmitting and receiving mechanism 20, which transmits the signal toward the objects 12, 14. The signal reflected by the object is received by the transmitting and receiving mechanism 20 and mixed with a portion of the transmit signal in the mixer 18. A baseband signal b is thus obtained, which is further evaluated in an electronic evaluation and control mechanism 22. The control and evaluation mechanism 22 includes a control unit 24 that controls the function of the oscillator 16. In radar measurements, the frequency of the transmit signal provided by the oscillator is modulated by a sequence of rising or falling ramps.

[0028] 2 shows the transmission frequency f of the transmission signal 28 as a function of time t. During the measurement, the transmitting antenna transmits two sequences of lamps with identical lamp parameters, which are interleaved in time with each other. A first sequence 30 of lamps 32 is shown in FIG. 2 with a solid line, while a second sequence 34 of lamps 36 is shown with a dashed line. The number i of the sequence to which the lamp belongs and the respective lamp index j of the lamp within a sequence are provided.

[0029] Each ramp 36 of the second sequence 34 is offset in time T with respect to the ramp 32 of the first sequence 30 having the same ramp index j. 12 Within each sequence 30, 34, successive lamps 32 or 36 are offset by a time interval T r2r That is, the time interval T r2r is the same for both sequences. Furthermore, there is one pause P between each two successive ramps of one sequence.

[0030] In the example shown in Figure 2, the difference in lamp center frequencies between successive lamps 32 or 36 in a sequence 30, 34 is equal to zero, so all lamps have the same frequency profile. The lamp center frequency here corresponds to the average transmission frequency f0.

[0031] FIG. 5 shows, in a schematic manner, a more detailed block diagram of the determination of the relative speed carried out by the control and evaluation unit 22 . The received and sampled baseband signals b1 and b2 obtained for the respective sequences of lamps 30, 34 are each subjected to a two-dimensional Fourier transform (2D-FFT). The first dimension corresponds to the transformation of the baseband signal obtained for each lamp. The second dimension corresponds to the transformation for the sequence of lamps, i.e., for the lamp index j. It is preferable that the amount of each transformation, i.e., the respective number of bins (sampling points or nodes), is consistent for all spectra in the first dimension and consistent for all spectra in the second dimension.

[0032] The relative velocity of the radar target 12 and the time delay T between the individual sequences of the ramps 12 Based on the partial measurements corresponding to the respective phases of the two partial measurements, a phase difference between the two partial measurements is generated. This phase difference is obtained as the phase difference between the complex amplitudes (spectral values) of peaks occurring at the same position in both two-dimensional spectra 50, 52. However, due to the relatively large time shift T between the corresponding ramps 32, 36 of both sequences 30, 34, 12 Therefore, the determination of the phase difference of both partial measurements does not allow a direct inductive inference to the relative velocity, since the periodicity of the phase leads to an ambiguity of the value of the relative velocity that belongs to a single phase difference.

[0033] The resulting two-dimensional spectra 50 or 52 are coherently summed in a first function block 54 by a "velocity beamformer." The velocity beamformer allows for enhancing or suppressing important velocities (e.g., velocities near 0 m / s for ACC objects or velocities near the negative self-velocity of the vehicle for stationary targets). Examples of important velocities would be, for example, velocities near 0 m / s for ACC objects or velocities near the negative self-velocity of the vehicle for stationary targets, or even the velocities of targets recognized as particularly important during object tracking. For beamforming, the velocity beamformer may utilize one of the many beamforming methods known from the literature, such as delay-and-sum or conventional beamforming, minimum variance distortionless response (MVDR) or Capon beamforming, or null steering (e.g., H.L. Van Trees, "Optimum array processing - Part IV of detection, estimation, and modulation theory", John Wiley, 2002; J. Capon, "High-Resolution Frequency-Wavenumber Spectrum Analysis", Proceedings of the IEEE, 1969, Vol. 57, pp. 1408-1418; or H. Cox, R. Zeskind, and M. Owen, "Robust adaptive beamforming", IEEE Transactions on Acoustics, Speech and Signal Processing, vol. 57, pp. 1408-1418). Processing, vol. 35, no. 10, pp. 1365-1376, 1987).

[0034] The domain of each two-dimensional spectrum 50, 52 is a two-dimensional raster of interval-velocity bins (dv bins), each assigned a specific value of the interval variable d and velocity variable v of one object. The spectral range is a set of complex numbers, which represent the complex amplitude of the Fourier transformed signal for each dv bin. In the beamforming methods described above, the spectral values, i.e., the complex amplitudes in each bin, are generally multiplied by a beamforming function, which depends on the velocity variable v and the time offset of the sequence. The form of this beamforming function varies depending on the beamforming method applied.

[0035] For example, in the case of the delay-and-sum method, for a rate variable v and N sequences (N=2 in the example based on Figures 2 and 5), a i (v)=(1 / √N)exp(2πj(2 / c)f0t i v) (1) where j is the imaginary unit, i is a subscript counting the sequence (i=1...N), f0 is the center frequency of the lamp, π is the ratio of the circumference of a circle to its diameter, and t i (=(i-1(T 12 )) is the time offset between the first lamp and the ith lamp (in the example based on Figs. 2 and 5, i only has values ​​1 and 2, and t1 = 0 and t2 = T 12 holds)) with a vector beamforming function a i (v) is used.

[0036] The spectra, modified by multiplication with the beamforming function, are then coherently summed for sequence index i, resulting in an N-fold increase in the signal-to-noise ratio.

[0037] As a further example, a MVDR beamforming method is briefly outlined, which can target and reduce sensitivity to a particular velocity v0. In this method, d(v,v0)=(R -1 (v0)a(v)) / (a H(v)R -1 (v0)a(v)) R(v0)=a(v0)a H (v0)+σI (2) where R(v0) is the correlation matrix, a(v0) is a vector whose components are defined as in equation (1), and a H (v) is the Hermitian adjoint vector of a(v). I is the identity matrix of dimension NxN. The term σI serves to regularize the correlation matrix R(v) and ensures that the matrix R(v) can be inverted. In this regard, a beamforming function w(v, v) with σ can be used (σ can be chosen to avoid noise enhancement).

[0038] In the case of angle-resolved radar sensors with multiple transmit antennas, beamforming can be applied to the received signals of each transmit antenna, followed by non-coherent averaging over the transmit antennas (coherent summation is not possible here, since the phase also depends on the object's positioning angle).

[0039] Non-coherent summation results in a two-dimensional power spectrum 56. Then, in function block 58, target detection is performed by known methods, for example CFAR (Constant False Alarm Probability).

[0040] The location of the peak corresponding to the radar target 12 in the power spectrum 56, presented below as bin k, l, corresponds to the location of the peak in the individual spectra 50, 52. From the first dimension, correspondingly, bin k of the peak location, a linear relationship between the radar target relative velocity v and separation d is obtained based on the FMCW equation k=2 / c(d F + f v T), where c is the speed of light, F is the ramp stroke, T is the ramp duration of a single ramp 32 or 36, and f is the average transmit frequency.

[0041] Figure 3 shows a schematic graph of relative velocity v versus distance d. The linear relationship between v and d is plotted as a straight line. The Doppler frequency resulting from relative movement at velocity v varies over a relatively large time interval T r2r Therefore, in the illustrated example, the time interval T r2r Based on this, the information about the relative velocity of the radar target obtained from the Doppler frequency sampling has ambiguity based on a predetermined interval. In addition to the line vd resulting from frequency bin k, the periodic values ​​of the relative velocity v determined from frequency bin l are shown by dashed lines. The intersections with the line vd are marked. These intersections correspond to the possible pairs (v, d) of relative velocity and separation of the detected radar target 12. The actual targets for which the velocity v is to be determined are marked with a cross in FIG. 3.

[0042] From this, the ambiguity of the determined velocity v is resolved as explained below: The information v* about the periodic value of the relative velocity v under consideration is passed to a second function block 60 (FIG. 5), which also obtains complex two-dimensional spectra 50, 52 of the partial measurements.

[0043] To evaluate the measured phase difference, a control vector of the ideal measurement is calculated, which depends on the relative velocity v. The components of this control vector are the same as in equation (1). A measurement vector is then formed, which differs from the control vector by its components being formed by complex amplitudes (spectral values) according to the actual measured values, rather than by predicted velocity-dependent complex values. The control and measurement vectors are then multiplied by a scalar to determine the degree of agreement. Normalization of this product then yields a likelihood function.

[0044] Figure 4 shows, schematically, the relative velocity spectrum S(v) as a function of the relative velocity v, using a solid sinusoidal line. The maxima of the likelihood function correspond to the most probable values ​​of the parameter v. The relative velocity spectrum S(v) is inherently ambiguous, i.e., each maximum at a maximum value of 1 corresponds to the best match between the ideal phase shift occurring for that relative velocity v and the measured phase shift based on the measurement vector.

[0045] However, the function S(v) only needs to be evaluated at points 40 corresponding to periodic values ​​of the relative velocity v obtained from the evaluation based on the position of the peaks in bin (k,l). These points 40, corresponding to the velocity values ​​of the intersections marked in Figure 3, are marked on the curved trajectory of the function S(v) in Figure 4. In the illustrated example, the maximum agreement occurs at relative velocity v = 0 m / s, where the function S(v) has a predicted maximum value of 1, which corresponds to the actual value of the relative velocity v.

[0046] This allows the ambiguity arising from the position of the peak to be resolved with the additional information from the phase relationship. Based on the linear relationship, an estimate of the spacing d attributable to a selected estimate of the relative velocity v is determined.

[0047] A second function block 60 outputs determined estimates of the relative velocity v and the separation d. The essential steps of the method are shown as a flow diagram in Figure 6. In steps S1 to S7, shown as blocks in Figure 6, the following actions are performed: S1 2D-FFT for each ramp sequence (and each receive channel) Coherent addition of the individual spectra of ramp sequences for each transmit and receive antenna with S2-rate beamforming S3 Non-coherent or coherent addition of spectra for transmit and / or receive antennas S4 Target detection using CFAR with peak interpolation S5 Power comparison and possibly integration of spectra for the detection list and for all detections S6 Resolution of velocity ambiguity and overlap and possibly transmitter allocation (in case of Doppler division multiplexing) S7 Angle estimation by phase correction A comparison of the spectral power before and after velocity beamforming or between different velocity beamformers allows determining whether there is only one stationary target in this spectral cell (large power difference due to beamforming) or whether there are possibly more targets present (smaller power difference). This information can be used in three different ways: - Probability of one stationary target; - (additional) criteria for applying multi-target models during velocity disambiguation and for separating the signal contributions of stationary and moving targets for subsequent angle estimation; - Improved accuracy of peak interpolation allows for more accurate velocity based on multi-target models.

[0048] The effect achieved by velocity beamforming is shown graphically in Figures 7-9, each plotting the relative received power Q against the relative velocity of the positioned object. The thick curves 62 each show the results of the delay-and-sum method, while the thinner curves show the results of the MVDR method.

[0049] In Figure 8, the range of relative velocities around the value zero is shown with higher resolution, where it can be seen that the relative received power, and therefore the sensitivity of the sensor, clearly increases at zero relative velocity.

[0050] In the MVDR method, parameters were chosen to reduce sensitivity at relative velocities around -30 m / s (stationary targets when the vehicle's own velocity is 30 m / s). Figure 9 shows this velocity range with higher resolution. The pronounced minimum in curve 64 at v=-30 m / s can be clearly seen. This suppression of stationary targets clearly improves target detection, and spectral power comparisons aid in disambiguation in steps S4-S6. [Explanation of symbols]

[0051] 10 FMCW radar sensors 12 Radar Target 14 Objects 16 Voltage Controlled Oscillator 18 Mixer 20 Transmission and Reception Mechanisms 22 Control and Evaluation Mechanisms 24 Control Unit 28 Transmitted Signal 30 First Sequence 32 Lamp 34 Further Sequences 36 Lamp 40 Points corresponding to periodic values ​​of relative velocity v 50;52 Two-dimensional spectrum 54 First Function Block 56 Power Spectrum 58 Function Blocks 60 Second Function Block 62 Curve showing the results of the delay-and-sum method 64 Curve showing the results of the MVDR method b1;b2 baseband signals d. Vehicle spacing f Transmission frequency f0 average transmission frequency j Lamp index i The number of the sequence to which the lamp belongs k,l peak position T r2r Time interval T 12 Time lag v relative velocity

Claims

1. 1. A method for determining the relative velocity (v) of a radar target (12), comprising: (a) FMCW radar measurement is performed, in which a ramp-like frequency modulated transmission signal (28) is transmitted, and the modulation pattern of the transmission signal (28) changes over a certain time interval (T ｒ２ｒ ) and a first sequence (30) of successive lamps (32) spaced apart in time by the same time interval (T ｒ２ｒ and at least one further sequence (34) of successive lamps (36) spaced apart in time, said sequences (30; 34) being interleaved in time with one another; (b) The received signal is a baseband signal (b １ ;b ２ ) and the baseband signal (b １ ;b ２ ), a two-dimensional spectrum (50; 52) is calculated for each of said sequences (30; 34) separately by a two-dimensional Fourier transform, in a first dimension for each lamp and in a second dimension for the lamp index (j), (c) the baseband signal (b １ ;b ２ 1. A method for determining a value of the relative velocity of a radar target (12) based on the positions (k, l) of peaks in at least two of the two-dimensional spectra (50; 52) of the radar target (12), comprising: Each of the two-dimensional spectra (50; 52) formed for the different sequences (30; 34) contains a velocity variable (v) and a time shift (T) of the sequence of assignment to the first one of the sequences (30). １２ ), and the relative velocity of the radar target (12) is determined based on a coherent sum of the transformed spectra.

2. The method of claim 1 , wherein the beamforming function is formed according to a delay-and-sum method.

3. The method of claim 1 , wherein the beamforming function is formed according to the MVDR method.

4. 2. The method of claim 1, wherein for a radar sensor with multiple transmitting antennas and / or multiple receiving antennas, after coherent addition of the transformed spectra, non-coherent or coherent addition is performed for the transmitting and / or receiving antennas, and detection of individual radar targets is performed based on the result of the coherent addition.

5. The time interval (T) with successive lamps (32, 36) in a single sequence (30, 34) ｒ２ｒ 2. The method of claim 1, wherein the first dimension is selected such that undersampling occurs in the second dimension and the resulting disambiguation is performed based on the transformed spectrum.

6. The method of claim 5 , wherein a comparison of the spectral power in the modified and unmodified spectra distinguishes between single target and multiple target scenarios.

7. 2. The method of claim 1, wherein the beamforming function is selected such that, for a radar sensor moving at an ego-velocity relative to its surroundings, the radar sensor is desensitized to targets whose relative velocity is the inverse of the ego-velocity.

8. The method of claim 4 , wherein the beamforming function is selected to increase the sensitivity of the radar sensor to targets with zero relative velocity.

9. 9. An FMCW radar sensor comprising a control and evaluation mechanism (22) in which the method according to any one of claims 1 to 8 is implemented.

Images