METHOD FOR TREATING RADAR SIGNALS OF A RADAR SYSTEM AND RADAR SYSTEM

Abstract

Method for processing radar signals from a radar system having a plurality of antennas (330), the method comprising: • processing a plurality of radar signals (xm(k)) to determine a distance between the radar system (330) and at least one target and a speed of the at least one target, thereby forming a plurality of processed radar signals (DFTN), wherein each radar signal of the plurality of radar signals (xm(k)) is received from an associated antenna of the plurality of antennas (330) (410); • digitally beamshaping the plurality of processed radar signals (DFTN) for at least one beam direction, thereby forming a plurality of beam-shaped radar signals (420); and • summing the plurality of beam-shaped radar signals from the plurality of antennas per beam direction (430);• where the summation includes both coherent summation and non-coherent summation of the majority of beam-shaped radar signals from the majority of antennas per beam direction.;

Description

Technical field Various embodiments generally refer to a method for handling radar signals of a radar system and to a radar system. background A state-of-the-art radar signal processing method can involve beam shaping of a radar signal. Several beam shaping methods exist; one of them is digital beam shaping, which is very well established. To shape the beam of a radar signal, a portion of the signal coming from a specific direction can be digitally amplified. This may require two operations. In a first operation, the signal can be directed in a specific direction, and in a second operation, the signal can be amplified. The generation of the beam-shaped signal may require two passes through a signal processing unit, where the first pass can perform beam control and a further pass can process the signal with an uncontrolled beam. This process, especially the double pass, can be time-consuming and suboptimal. German patent application DE 10 2018 127 947 B3 discloses a method for detecting radio targets, which includes providing a digital radar signal with a sequence of signal segments (chirps) and detecting one or more radar targets based on a first subsequence of successive signal segments of the sequence. For each detected radar target, a distance value and a velocity value are determined. If a group of radar targets with overlapping signal components is detected, a spectral value is calculated for each radar target in the group based on a second subsequence of the signal segment sequence and further based on the velocity values ​​determined for the group of radar targets. DE 196 51 540 A1 discloses a frequency-modulated CW radar which uses a transmitter and two receivers that supply beat signals to an analyzer where they are digitized for phase calculation, differentiation and comparison. DE 10 2014 212 284 A1 discloses a MIMO-FMCW radar sensor for locating a radar target, in which an FMCW radar measurement is performed with a transmit signal whose modulation pattern for different transmit switching states, which differ in the selection of the antenna elements used for transmitting, comprises temporally interleaved sequences of ramps, from a position of a peak in a two-dimensional spectrum ambiguous values ​​for the relative velocity of the radar target. Summary A method for processing radar signals according to claim 1 and a radar system according to claim 16 are provided. Further embodiments are described in the dependent claims. A method for processing radar signals from a radar system with a plurality of antennas is provided. The method may include processing a plurality of radar signals to determine a distance between the radar system and at least one target and a velocity of the at least one target, thereby generating a plurality of processed radar signals, wherein each radar signal of the plurality of radar signals is received by an associated antenna of the plurality of antennas, digitally beamshaping the plurality of processed radar signals for at least one beam direction, thereby generating a plurality of beam-shaped radar signals, and summing the plurality of beam-shaped radar signals from the plurality of antennas for each beam direction. Brief description of the drawings In the drawings, identical reference numerals generally refer to the same parts in the different views. The drawings are not necessarily to scale; instead, the emphasis is generally on illustrating the principles of the invention. The following description describes various embodiments of the invention with reference to the following drawings, in which: Fig. 1A shows an exemplary arrangement of vehicles to illustrate how radar signals are processed, and shows the processed radar signals; Fig. 1B shows beam-guided radar signals; Fig. 1C schematically shows part of a radar system according to a prior art; Fig. 1D schematically shows part of a radar system according to a prior art; Fig. 2A schematically shows part of a radar system according to various embodiments; Fig.Figure 2B schematically shows a part of a radar system according to different embodiments; Figure 3 schematically shows a radar system according to different embodiments; and Figure 4 shows a flowchart of a method for handling radar signals of a radar system according to different embodiments. Description The following detailed description refers to the accompanying drawings, which show certain details and embodiments in which the invention can be implemented. The word "exemplary" is used here in the sense of "serving as an example or illustration". Each embodiment or design described here as "exemplary" is not necessarily to be understood as preferred or advantageous over other embodiments or designs. Various aspects of the disclosure are provided for devices, and various aspects of the disclosure are provided for methods. It is understood that fundamental properties of the devices also apply to the methods and vice versa. Therefore, for the sake of brevity, a duplicate description of such properties may have been omitted. Fig. 1A shows, in an upper view, an exemplary arrangement of targets T1 and T2 (in this case, cars) to be detected by a radar system operated in an object (e.g., another car) labeled RS, to illustrate how radar signals from the radar system can be processed, e.g., in a part 100 of the radar system, as shown in Fig. 1C. A lower view shows the processed radar signals for the first target T1 and for the second target T2 as normalized power as a function of range (where range is the distance between the radar system and the respective target). The radar signals shown in Fig. 1A can be unguided, which may be similar to or identical with a beam-guided radar signal for an angle of 0° (in other words, straight ahead in front of object RS). Fig. 1B shows beam-guided radar signals for five different beam guidance angles: -40°, -25°, 0°, 25°, and 40°. These five angles are also shown in Fig. 1A. As can be seen with the beam-guided radar signals, both targets T1 and T2 are detected in all beam-guided signals, but the relative signal strengths for the two targets T1 and T2 can vary. While for most beam guidance angles the normalized power of the radar signal reflected from the first target T1, which is the closer of the two targets T1 and T2, may be higher than that for T2, it may reach its highest absolute value for the beam guidance angle corresponding to the angle of 0° between the radar system RS and the target T1. For the beam guidance angle of -25°, which corresponds to the direction to the second target T2, the signal strength of the second target T2 reaches its highest value, which is approximately the same as that of the first target T1.This can enable the determination of directions to the first target T1 and to the second target T2. In simpler terms, digital beam control can involve combining radar signals received by a plurality of receivers in such a way as to take into account a relative phase delay that would be caused in the majority of receivers by a target located in the direction of a desired control angle. Fig. 1C schematically shows a part 100 of the radar system which may be configured to apply digital beam control according to a prior art. The radar system can be configured to coherently integrate radar signals from a plurality of antennas provided by an input array 102. The coherent integration is shown in the upper branch and can include beam shaping 104, complex windowing 106, a (fast) Fourier transform 108, coherent summation 110 of the beam-shaped, windowed, and Fourier-transformed signals from the plurality of antennas (also called sum-over-antennas), and subsequent amplitude sensing 112 (labeled as "power"). The radar system can further be configured to non-coherently integrate the radar signals from a plurality of antennas provided by the input group 102. The non-coherent integration is shown in the lower branch and can include complex windowing 114, a (fast) Fourier transform 116, amplitude sensing 118 (referred to as "power"; since this is applied before summing, phase information is lost, making the sum non-coherent), and non-coherent summing 120 of the absolute values ​​of the windowed and Fourier-transformed signals from the plurality of antennas (also referred to as sum-over-antennas). The coherent sum and the non-coherent sum can be forwarded to a capture block 122 for target acquisition, e.g., a capture block 122 that is set up to capture a constant false alarm rate (CFAR) of the processed signal and to determine if and where the processed signal exceeds the constant false alarm rate. The beam shaping described above may require two passes in a signal processing unit: one in which the beam is controlled (the coherent integration in the upper branch), and another in which the beam remains uncontrolled (the non-coherent integration in the lower branch). Fig. 1D schematically shows further details of part 101 of part 100 of the radar system from Fig. 1C. In particular, it shows how the beam shaping 104 and the complex windowing 106 can be carried out, and a formula is provided for calculating the digital beam-shaped Fourier transform for a given beam steering direction (specified as beam index b) DBFb(n). Descriptions of the variables / labels in the formula are given in Table 1 below. In particular, a single complex multiplier 134 can be provided during a first mathematical operation (MATH1). This multiplier includes both a windowing coefficient w(k), which can differ for different samples but be the same for different beam steering angles and antennas, and a weighting coefficient, also called the antenna factor a(b,m), which can differ for different beam directions b and antennas m but be the same for different samples. The multiplier 134 can thus be called the "multiplied antenna factor with windowing" (a(b,m)·w(k)). The multiplier 134 can be stored in memory, for example, in a configuration RAM. For each desired beam control direction with beam index b, this may require storing M*K (for M antennas and K samples per antenna). Therefore, a large amount of memory may be required. For example, for M = 8, K = 1024, a complex 32-bit configuration RAM size of 64 KiB may be required, which can exceed the size of a typical configuration RAM. Furthermore, the M*K elements can be updated between successive loops targeting different beam directions (different beam indices b), e.g. by a CPU, which can cause high update latency. According to various embodiments, a radar system and a method for processing radar signals from a radar system are provided, which support the use of a complex multiplication factor per antenna for coherent integration or for non-coherent integration of the radar signal. In various embodiments, a fixed digital beamforming coefficient (DBF) can be provided per antenna to enable beam control per antenna. In various embodiments, finer beam shaping, e.g. beam control per sample, can be enabled by providing a vector coefficient. In various embodiments, the windowing coefficient w(k) and the antenna factor a(b,m) can be decoupled. This can enable beam steering to be performed after the Fourier transform, in other words, post-FFT beam steering, which means that beam steering can be performed in parallel. In one exemplary embodiment, this can result in a saving of up to 500 µs (from 2.5 ms) per processing step. Fig. 3 schematically shows a radar system 300 in accordance with various embodiments, Fig. 2A schematically shows a part 200 of a radar system in accordance with various embodiments, for example of the radar system 300 from Fig. 3, and Fig. 2B schematically shows a part 201 of the part 200 of the radar system from Fig. 2A. The following table describes the terms used in the formulas in Fig. 1D, Fig. 2B, and elsewhere in this document. Table 1 Number of antennas m = {0,1, ..., (M - 1)} Antenna index Number of samples per antenna k = {0,1, ..., (K - 1)} Sample index (time domain) xm(k)Sample of antenna-m w(k)Window function Xm(n) = DFTN(w(k) · xm(k))N-point DFT of antenna-mn = frequency range-sample index= {0,1 ..., (N-1)} Number of beam shaping directions b = {0,1, ..., (D - 1)} Radiation index ab,m Antenna factor, complex value. Digital beamforming example formula for calculating the coefficient: ab,m = exp(2πj(m · d) · sin(θb)) θb beam direction (in radians) d antenna spacing (in wavelength) The radar system 300 can have a plurality of antennas 330 (for example, M antennas, addressed as m=0,...M-1) and at least one processor 332 configured to process a plurality of radar signals xm(k) to determine the distance between the radar system 300 and at least one target and the velocity of the at least one target. Figure 1A can also be considered here to illustrate the general operating principle, geometry, etc. In various embodiments, a plurality of processed radar signals DFTN can be generated, wherein each radar signal of the plurality of radar signals can be received by an associated antenna of the plurality of antennas 330. In other words, one radar signal can be generated per antenna 330. In various embodiments, a plurality of sample values ​​(e.g., K sample values, addressed as k=0,...K-1) can be generated per antenna 330. The processor 332 can further be configured to perform digital beamshaping (see, for example, 204 in Fig. 2A and Fig. 2B) of the majority of the processed radar signals DFTN for at least one beam direction b, optionally for a plurality of beam directions (e.g., D beam directions, addressed as b=0,...D-1), thereby forming a plurality of beam-shaped radar signals, and to summate the plurality of beam-shaped radar signals from the plurality of antennas 330 per beam direction b (see, for example, 210 in Fig. 2A and Fig. 2B). One result can be a Fourier transform of the digitally beam-shaped radar signals for beam direction b (or for each of the plurality of beam directions b) DBFb(n). The processing of the majority of radar signals xm(k) can include at least one Fourier transform 208, e.g., an N-point Fourier transform, and / or a windowing process 206. The Fourier transform and / or the windowing can be used to determine the distance and / or Doppler velocity of the target. See Fig. 2A and Fig. 2B for an illustration. The Fourier transform and / or windowing can be performed essentially as is known in the art. For example, the windowing can involve multiplying each of the samples by a sample-specific complex windowing coefficient w(k). The windowing can be performed in various embodiments prior to the Fourier transform. The digital beamforms 204 of the majority of processed radar signals can exhibit the multiplication of the processed radar signal of each of the antennas 330 with an antenna- and beam-specific complex multiplication factor. In various embodiments, the antenna- and beam-specific complex multiplication factors can be provided in a matrix that has a number M of the majority of the antennas as a first dimension and a number D of the majority of the beam directions as a second dimension. The summing 210 exhibits coherent summing of the majority of the beam-shaped radar signals from the majority of the antennas 330 per beam direction b. In various embodiments, the summed beam-shaped radar signal for the at least one beam direction (b) can be formed according to the formula: where b is a beam direction index, m is an antenna index, ab is an antenna- and beam-specific complex multiplication factor, where k is a sample index, Xm(n) is the Fourier transform (DFT) of a multiplication between a complex window function wm(k) for an antenna m for the samples k and the radar signals xm(k) for antenna m for the sample values ​​k. In various embodiments, the beam shaping 204 can be performed on-the-fly from the windowed radar signal, which may be stored in a memory 236, e.g. a buffer, before the summation is carried out by performing the complex multiplication with the multiplication factor. The antenna- and beam-specific complex multiplication factor ab,m can be defined in various embodiments as or as where b = beam index, m = antenna index and hann() can be the Hann window function. The angle Θ, which indicates the beam direction that is to receive greater weight in the beam-shaped radar signal, can thus influence the antenna- and beam-specific complex multiplication factor ab,m. The coherent sum can be determined in various embodiments from: In various embodiments, the summing further comprises the non-coherent summation of the majority of the beam-shaped radar signals from the majority of the antennas 330 per beam direction Θ. In the exemplary embodiment shown in Fig. 2A, this is carried out in the lower branch. The beam shaping 216 can be performed on-the-fly from the windowed radar signal, which may be stored in a memory 236, e.g., a buffer, before the power is determined and summed. The non-coherent sum can be determined in various embodiments from: where cb,my is antenna- and beam-specific multiplication factor for the non-coherent sum, which may be identical to or different from the antenna- and beam-specific multiplication factor ab,m for the coherent sum. The antenna- and beam-specific multiplication factor for the non-coherent sum cb,m can be determined in various embodiments from or from where b = beam index, m = antenna index, and hann() is the Hann window function. The antenna- and beam-specific multiplication factor for the non-coherent sum cb,m can be determined similarly to ab,mund / or even as a combination with a window function suitable for the antenna index m. In various embodiments, the processor 332 can be configured to perform non-coherent summation and coherent summation in parallel. In various embodiments, an additional complex multiplier ab,m is provided (compared, for example, to the prior art in Fig. 1C and Fig. 1D), which is processed in the part designated “MATH2”. The additional complex multiplier ab,m can be generated by separating the antenna factor a(b,m) from the windowing w(k). To compare the required storage space, we take the same example as above (e.g., M=8 antennas, K=1024 samples, and a windowing w(k), real 32-bit), and 4 KiB may be required (i.e., only about 1 / 16 of the storage requirement according to the state of the art). The antenna factor a(b,m) may only require one byte of storage space per antenna, so a total of about 64 bytes. According to various embodiments, the antenna factor a(b,m) can be updated by the processor 332, e.g., a CPU, between different beam passes, i.e., between passes with different beam indices, or stored as a fixed configuration. The latter, in particular, can increase the execution speed, but due to the small size of the antenna factor a(b,m), updating the antenna factor a(b,m) can also be significantly faster than in the prior art. According to various embodiments, the radar system 300 can also have at least one register (e.g., 64 bits) per antenna 404. This at least one register can, for example, be contained in the part of the radar system 300 in which the non-coherent integration (NCI) is performed, also referred to as the NCI module. In various embodiments, both the coherently integrated radar signal and the non-coherently integrated radar signal can be fed to the detection block 122, which may be similar to or identical with a detection block as used in the prior art. The coherently integrated radar signal and the non-coherently integrated radar signal can be processed together to eliminate false detections, in other words, to identify "real" targets. Fig. 4 shows a flowchart 400 of a method for handling radar signals of a radar system in accordance with various embodiments. The method may include processing a plurality of radar signals to determine a distance between the radar system and at least one target and a speed of the at least one target, thereby forming a plurality of processed radar signals, wherein each radar signal of the plurality of radar signals is received by an associated antenna of the plurality of antennas (in 410), digitally beam-shaping the plurality of processed radar signals for at least one beam direction, thereby forming a plurality of beam-shaped radar signals (in 420), and summing the plurality of beam-shaped radar signals from the plurality of antennas per beam direction (in 430). The following are several examples: Example 1 is a method for processing radar signals from a radar system with a plurality of antennas. The method may include processing a plurality of radar signals to determine a distance between the radar system and at least one target and a speed of the at least one target, thereby generating a plurality of processed radar signals, wherein each radar signal of the plurality of radar signals is received by an associated antenna of the plurality of antennas, digitally beamshaping the plurality of processed radar signals for at least one beam direction, thereby generating a plurality of beam-shaped radar signals, and summing the plurality of beam-shaped radar signals from the plurality of antennas for each beam direction. In Example 2, the subject of Example 1 may optionally exhibit that the processing of the majority of radar signals involves a Fourier transformation. In Example 3, the subject of Example 1 or 2 may optionally feature windowing in the processing of the majority of radar signals. In Example 4, the subject of Example 3 may optionally feature that each radar signal of the plurality of radar signals has a plurality of samples, and that the windowing features the multiplication of each of the samples by a sample-specific complex windowing coefficient. In Example 5, the subject of Examples 2 and 3 can optionally exhibit that windowing is applied before the Fourier transform. In Example 6, the object of one of Examples 1 to 5 may optionally have that at least one beam direction has a plurality of beam directions. In Example 7, the subject of one of Examples 1 to 6 may optionally feature that the digital beamforming of the majority of processed radar signals involves the multiplication of the processed radar signal of each of the antennas by an antenna- and beam-specific complex multiplication factor. In Example 8, the subject of Example 7 can optionally feature that the antenna- and beam-specific complex multiplication factors are provided as a matrix having a number of the plurality of antennas as a first dimension and a number of the plurality of beam directions as a second dimension. In Example 9, the subject of one of Examples 1 to 8 exhibits that the summation is a coherent summation of the majority of beam-shaped radar signals from the majority of antennas per beam direction. In Example 10, the subject of one of Examples 1 to 9 may optionally feature that the summed beam-shaped radar signal for the at least one beam direction (b) is derived from the formula: where b is a beam direction index, m is an antenna index, ab is my antenna- and beam-specific complex multiplication factor; where k is a sample index, Xm(n) is the Fourier transform (DFT) of a multiplication between a complex window function wm(k) for an antenna m for the samples k and the radar signals xm(k) for the antenna m for the samples k, and where, optionally, or where b = m and hann() is the Hann window function. In Example 11, the subject of Examples 9 and 10 may optionally exhibit that it is the coherent sum. In Example 12, the subject of one of Examples 9 to 11 exhibits that the summation further involves a non-coherent summation of the majority of beam-shaped radar signals from the majority of antennas per beam direction. In Example 13, the subject of Example 12 may optionally have that the non-coherent sum is , where cb is my antenna- and beam-specific multiplication factor for the non-coherent sum, and that, optionally, or where b = m and hann() is the Hann window function. In Example 14, the subject of one of Examples 9 to 11 and Example 12 or 13 may optionally feature that the non-coherent summation and the coherent summation are performed in parallel. In Example 15, the subject of one of Examples 9 to 11 may optionally include the coherent sum being provided to a target acquisition algorithm for performing a target acquisition and / or to a local maxima search algorithm for searching for local maxima. In Example 16, the subject of Example 12 or 13 may optionally include the non-coherent sum being provided to a target acquisition algorithm for performing a target acquisition and / or to an algorithm for searching for local maxima. In Example 17, the subject of Example 12 or 13 may optionally feature that target acquisition and the search for local maxima are performed in parallel. Example 18 is a radar system. The radar system can have a plurality of antennas and at least one processor configured to process a plurality of radar signals in order to determine a distance between the radar system and at least one target and a speed of the at least one target, thereby forming a plurality of processed radar signals, wherein each radar signal of the plurality of radar signals is received by an associated antenna of the plurality of antennas, the plurality of processed radar signals are digitally beamshaped for at least one beam direction, thereby forming a plurality of beam-shaped radar signals, and the plurality of beam-shaped radar signals from the plurality of antennas are summed for each beam direction. In Example 19, the subject of Example 18 may optionally feature that the processing of the majority of radar signals involves a Fourier transform. In Example 20, the subject of Example 18 or 19 may optionally feature that the processing of the majority of radar signals includes windowing. In Example 21, the subject of Example 20 may optionally feature that each radar signal of the plurality of radar signals contains a plurality of samples, and that the windowing involves multiplying each of the samples by a sample-specific complex windowing coefficient. In Example 22, the subject of Examples 20 and 21 may optionally include the processor being further configured to apply windowing before the Fourier transform. In Example 23, the object of one of Examples 18 to 22 may optionally have that at least one beam direction has a plurality of beam directions. In Example 24, the subject of one of Examples 18 to 23 may optionally feature that the digital beamforming of the majority of processed radar signals involves multiplying the processed radar signal of each of the antennas by an antenna- and beam-specific complex multiplication factor. In Example 25, the subject of Example 24 may optionally feature that the antenna- and beam-specific complex multiplication factors are provided as a matrix having a number of the plurality of antennas as a first dimension and a number of the plurality of beam directions as a second dimension. In Example 26, the subject of one of Examples 18 to 25 exhibits that the summation is a coherent summation of the majority of the beam-shaped radar signals from the majority of the antennas per beam direction. In Example 27, the subject of one of Examples 18 to 26 may optionally include that the processor is further configured to derive the summed beam-shaped radar signal for the at least one beam direction (b) from the formula: where b is a beam direction index, m is an antenna index, ab is my antenna- and beam-specific complex multiplication factor, where k is a sample index, Xm(n) is the Fourier transform (DFT) of a multiplication between a complex window function wm(k) for an antenna m for the samples k and the radar signals xm(k) for the antenna m for the samples k, and where, optionally, or where b = m and hann() is the Hann window function. In Example 28, the subject of Examples 26 and 27 may optionally exhibit that it is the coherent sum. In Example 29, the subject of Example 27 or 28 further indicates that the summation includes a non-coherent summation of the majority of beam-shaped radar signals from the majority of antennas per beam direction. In Example 30, the subject of Example 29 can optionally have that the incoherent sum is where cb,m is an antenna- and beam-specific multiplication factor for the incoherent sum, and where, optionally, or where b = m and hann() is the Hann window function. In Example 30, the subject of Example 26 to 28 and Example 29 or 30 may optionally include the processor further configured to perform non-coherent summation and coherent summation in parallel. While the invention has been shown and described, particularly with reference to specific embodiments, it should be clear to the person skilled in the art that various modifications in form and detail can be made without departing from the spirit and scope of the invention as defined by the accompanying claims. The scope of the invention is thus specified by the accompanying claims, and all modifications that fall within the scope and equivalence of the claims are therefore intended to be included.

Claims

Method for processing radar signals from a radar system having a plurality of antennas (330), the method comprising: • processing a plurality of radar signals (xm(k)) to determine a distance between the radar system (330) and at least one target and a speed of the at least one target, thereby forming a plurality of processed radar signals (DFTN), wherein each radar signal of the plurality of radar signals (xm(k)) is received from an associated antenna of the plurality of antennas (330) (410); • digitally beamshaping the plurality of processed radar signals (DFTN) for at least one beam direction, thereby forming a plurality of beam-shaped radar signals (420); and • summing the plurality of beam-shaped radar signals from the plurality of antennas per beam direction (430);• where the summation includes both coherent summation and non-coherent summation of the majority of beam-shaped radar signals from the majority of antennas per beam direction.; Method according to claim 1, wherein the processing of the plurality of radar signals (xm(k)) comprises a Fourier transformation (208). Method according to claim 1 or 2, wherein the processing of the plurality of radar signals (xm(k)) comprises a windowing (206). Method according to claim 3, wherein each radar signal of the plurality of radar signals (xm(k)) comprises a plurality of samples; and wherein the windowing comprises multiplying each of the samples by a sample-specific complex windowing coefficient (w(k)). Method according to claims 2 and 3, wherein the windowing is applied before the Fourier transform. Method according to any one of claims 1 to 5, wherein the at least one beam direction has a plurality of beam directions. Method according to any one of claims 1 to 6, wherein the digital beam shaping of the plurality of processed radar signals comprises multiplying the processed radar signal of each of the antennas by an antenna- and beam-specific complex multiplication factor (ab,m). Method according to claim 7, wherein the antenna- and beam-specific complex multiplication factors (ab,m) are provided as a matrix comprising a number of the plurality of antennas as a first dimension and a number of the plurality of beam directions as a second dimension. Method according to any one of claims 1 to 8, wherein the summed beam-shaped radar signal for the at least one beam direction (b) is given by the formula: DBF b ( n ) = ∑ m = 0 M − 1 ab , m X m ( n ) is derived, where b is a beam direction index, m is an antenna index, a b,m an antenna- and beam-specific complex multiplication factor; where X m ( n ) = DFTN ( wm ( k ) ⋅ xm ( k ) ) = ∑ k = 0 N − 1 wm ( k ) ⋅ xm ( k ) ⋅ e − i 2 π N kn where k is a sample index, X m (n) the Fourier transform (DFT) of a multiplication between a complex window function w m (k) for an antenna m for the samples k and the radar signals x m (k) for antenna m for the samples k; and where optional ab , m = exp ( 2 π j ( m ⋅ d ) ⋅ sin ( θ b ) ) or ab, m = exp (2 π j (m ⋅ d) ⋅ sin (θ b)) * hann (m), where b = beam index, m = antenna index and hann() is the Hann window function, θ b the beam direction (in radians) and d the antenna spacing (in wavelength). Method according to claim 9, wherein the coherent sum Coherent Sumb ( n ) = | DBF b ( n ) | 2 is. Method according to claim 9, wherein the non-coherent sum N on Coherent Sumb ( n ) = ∑ m = 0 M − 1 | cb , m X m ( n ) | 2 is, where c b,m an antenna- and beam-specific multiplication factor for the incoherent sum; and where, optionally, cb, m = exp (2 π j (m ⋅ d) ⋅ sin (θ b)) or cb, m = exp (2 π j (m ⋅ d) ⋅ sin (θ b)) * hann (m), where b = beam index, m = antenna index and hann() is the Hann window function, θ b the beam direction (in radians) and d the antenna spacing (in wavelength). Method according to any one of claims 1 to 11, wherein the non-coherent summation and the coherent summation are performed in parallel. Method according to any one of claims 1 to 10, wherein the coherent sum is provided to a target acquisition algorithm for performing a target acquisition and / or to an algorithm for searching for local maxima. Method according to claims 1 to 11, wherein the non-coherent sum is provided to a target acquisition algorithm for performing target acquisition and / or to an algorithm for searching for local maxima. Method according to claim 13 or 14, wherein target acquisition and the search for local maxima are performed in parallel. Radar system (300) comprising: • a plurality of antennas (330); and • at least one processor (332) configured to: ◯ process a plurality of radar signals (xm(k)) to determine a distance between the radar system (300) and at least one target and a velocity of the at least one target, thereby generating a plurality of processed radar signals (DFTN), each radar signal of the plurality of radar signals (xm(k)) being received from an associated antenna of the plurality of antennas (330); ◯ digital beamforms of the plurality of processed radar signals (DFTN) for at least one beam direction, thereby generating a plurality of beam-shaped radar signals; and ◯ summate the plurality of beam-shaped radar signals from the plurality of antennas (330) per beam direction;◯ where the summation includes both coherent summation and incoherent summation of the majority of beam-shaped radar signals from the majority of antennas (330) per beam direction.; Radar system (300) according to claim 16, wherein the processing of the plurality of radar signals (xm(k)) comprises a Fourier transformation (208). Radar system (300) according to claim 16 or 17, wherein the processing of the plurality of radar signals (xm(k)) comprises a windowing (206). Radar system (300) according to claim 18, wherein each radar signal of the plurality of radar signals (xm(k)) comprises a plurality of samples; and wherein the windowing comprises multiplying each of the samples by a sample-specific complex windowing coefficient (w(k)). Radar system (300) according to claims 18 and 19, wherein the processor (332) is further configured to apply windowing prior to Fourier transformation. Radar system (300) according to one of claims 16 to 20, wherein the at least one beam direction has a plurality of beam directions. Radar system (300) according to one of claims 16 to 21, wherein the digital beamforming of the plurality of processed radar signals (DFTN) comprises the multiplication of the processed radar signal of each of the antennas (330) by an antenna- and beam-specific complex multiplication factor (ab,m). Radar system (300) according to claim 22, wherein the antenna- and beam-specific complex multiplication factors (ab,m) are provided as a matrix comprising a number of the plurality of antennas (330) as a first dimension and a number of the plurality of beam directions as a second dimension. Radar system (300) according to one of claims 16 to 23, wherein the processor (332) is further configured to calculate the summed beam-shaped radar signal for the at least one beam direction (b) from the formula: DBF b ( n ) = ∑ m = 0 M − 1 ab , m DFTN ( w ( k ) xm ( k ) ) to derive, where b is a beam direction index, m is an antenna index, a b,m an antenna- and beam-specific complex multiplication factor; wherein X m ( n ) = DFTN ( wm ( k ) ⋅ xm ( k ) ) = ∑ k = 0 N − 1 wm ( k ) ⋅ xm ( k ) ⋅ e − i 2 π N kn where k is a sample index, X m (n) the Fourier transform (DFT) of a multiplication between a complex window function w m (k) for an antenna m for the samples k and the radar signals x m (k) for antenna m for the samples k is; and where, optionally, ab , m = exp ( 2 π j ( m ⋅ d ) ⋅ sin ( θ b ) ) or ab, m = exp (2 π j (m ⋅ d) ⋅ sin (θ b)) * hann (m), where b = beam index, m = antenna index and hann() is the Hann window function, θ b the beam direction (in radians) and d the antenna spacing (in wavelength). Radar system (300) according to claims 23 and 24, wherein the coherent sum Coherent Sumb ( n ) = | DBF b ( n ) | 2 is. Radar system (300) according to claim 24, wherein the non-coherent sum N on Coherent Sumb ( n ) = ∑ m = 0 M − 1 | cb , m X m ( n ) | 2 is, where c b,m an antenna- and beam-specific multiplication factor for the incoherent sum; and where, optionally, cb, m = exp (2 π j (m ⋅ d) ⋅ sin (θ b)) or cb, m = exp (2 π j (m ⋅ d) ⋅ sin (θ b)) * hann (m), where b = beam index, m = antenna index and hann() is the Hann window function, θ b the beam direction (in radians) and d the antenna spacing (in wavelength). Radar system (300) according to one of claims 16 to 26, wherein the processor (332) is further configured to perform the non-coherent summation and the coherent summation in parallel.

Images