RADAR SYSTEM AND PROCEDURES FOR OPERATING A RADAR SYSTEM

Abstract

The invention relates to a radar system, particularly a primary radar system, comprising at least one signal generating device (SGEN), which is configured to generate and to emit a transmit signal sequence, at least one signal detection device, which is configured to receive and to detect a receive signal sequence reflected on an object structure, at least one mixer (MIX) for mixing the receive signal sequence with the transmit signal sequence and for forming N baseband signals s.sub.b(n, t), where n=1 . . . N, and at least one scanning device (ADC), which is configured to scan the N baseband signals at scanning frequencies fs(n), wherein at least two, preferably at least three, further preferably all of the N scanning frequencies fs(n) differ from each other.

Claims

1. A radar system, particularly a primary radar system, comprising at least one signal generation device configured to generate and emit a transmit signal sequence, at least one signal detection device configured to receive and detect a receive signal sequence reflected on an object structure, at least one mixer (MIX) for mixing the receive signal sequence with the transmit signal sequence and for forming N baseband signals s.sub.b(n, t), with n=1 . . . N, and at least one scanning device configured to scan the N baseband signals at scanning rates f.sub.s(n), wherein at least two, preferably at least three, further preferably all, of the N scanning rates f.sub.s(n) differ from each other.

2. A radar system, particularly a primary radar system, preferably according to claim 1, comprising at least one signal generation device configured to generate and emit a transmit signal sequence, and at least one signal detection device configured to receive and detect a signal sequence reflected on an object structure, wherein the transmit signal sequence comprises a number of N frequency-modulated, preferably linear frequency-modulated, radar signals s.sub.Tx with a respective sweep rate value μ(n), with n=1 . . . N, wherein at least two, preferably at least three, further preferably all, amounts of the sweep rate values differ from each other.

3. The radar system according to claim 1 or 2, wherein the at least one scanning device is configured to implement the different scanning rates f.sub.s(n) by a different scanning clock directly in an analog-to-digital conversion.

4. The radar system according to claim 1, wherein the at least one scanning device is configured to implement the different scanning rates f.sub.s(n) algorithmically by means of a scanning rate conversion, particularly provided that the N baseband signals s.sub.b(n, t) were first scanned by an analog-to-digital converter (ADC) with a uniform scanning clock.

5. The radar system according to claim 1, wherein the at least one scanning device is configured to adjust at least two, preferably at least three, further preferably all, of the N scanning rates as a function of a modulation parameter—such as particularly a modulation speed, a sweep bandwidth, a sweep duration and/or a sweep rate—preferably such that these are proportional to the magnitude of the respective sweep rate μ(n).

6. The radar system according to claim 1, wherein the scanning device is configured such that scanning times t.sub.s(n, m) are selected such that the scanning points s.sub.b(n, m) with identical index m are associated with an identical carrier frequency value f.sub.c(m).

7. The radar system according to claim 1, wherein the signal generating device is configured such that: at least one of two edge frequencies (f.sub.ca(n), f.sub.ce(n)) of at least one sweep is equal to at least one of two edge frequencies of at least one other, preferably multiple other, further preferably all other, sweeps and/or lies within a sweep band of at least one other, preferably multiple other, further preferably all other, sweeps, and/or a sweep bandwidth B(n) is the same for at least two, preferably at least three, further preferably all, sweeps and/or sweep bands of at least two, preferably at least three, further preferably all sweeps at least overlap, particularly are identical.

8. The radar system according to claim 1, wherein the signal generating device is configured to generate at least one sweep rate value μ(n), particularly multiple or all sweep rate values μ(n) as random value(s) or pseudo-random value(s).

9. The radar system according to claim 1, comprising an evaluation device which is configured to form a number of M Doppler signals sb.sub.m(n, t), each having N scanning points, from scanning points s.sub.b(n, m) of the baseband signals s.sub.b(n, t), wherein a distance between at least two, preferably at least three, further preferably all, scanning points t.sub.s(n, m) and a respective adjacent scanning point is preferably different.

10. The radar system according to claim 1, in particular according to claim 8, wherein a respective scanning sequence is different for at least two or more, particularly all, Doppler signals, particularly depending on a selection of sweep edge times (t.sub.a(n), t.sub.e(n)).

11. The radar system according to claim 1, wherein the evaluation device is configured such that spectra Sb.sub.m(k) with K spectral values (k=1 to K) are calculated from the M Doppler signals M, particularly based on at least partially non-equidistantly scanned input values, wherein preferably multiple, particularly all M spectra at the same K discrete frequency base points k (k=1 to K) are calculated and, in a second step, a total of K signals Sb.sub.k(m) are formed from these M spectra Sb.sub.m(k), in each case from all scanning points with the identical index k, wherein further preferably these K signals Sb.sub.k(m) (with m=1, 2, . . . , M) are subjected to another spectral analysis and thus K spectra Sb.sub.k(j) (with j=1, 2, . . . , J) and/or a matrix with the values Sb(k, j) are formed, wherein the other spectral analysis is preferably a fast Fourier transformation (FFT), wherein further preferably a distance Doppler diagram is formed from this matrix with the values Sb(k, j), with which target objects are detected and to which a distance or speed is assigned.

12. A radar method, comprising: generating and sending a transmit signal sequence receiving and detecting a receive signal sequence reflected on an object structure, mixing the receive signal sequence with the transmit signal sequence and forming N baseband signals s.sub.b(n, t), with n=1 . . . N, and scanning the N baseband signals at scanning rates f.sub.s(n), wherein at least two, preferably at least three, further preferably all, of the N scanning rates f.sub.s(n) differ from each other.

13. A radar method, particularly a primary radar method, preferably according to claim 12, the method comprising: generating and transmitting a transmit signal sequence, wherein the transmit signal sequence comprises a number of N frequency-modulated, preferably linear frequency-modulated, radar signals s.sub.Tx with a respective sweep rate value μ(n), with n=1 . . . N, wherein at least two, preferably at least three, further preferably all, amounts of the sweep rate values differ from each other; receiving and detecting a receive signal sequence reflected on an object structure.

14. The radar method according to claim 12, comprising: forming a number of M Doppler signals sb.sub.m(n, t) with N scanning points each from scanning points s.sub.b(n, m) of the baseband signals s.sub.b(n, t), wherein a distance of at least two, preferably at least three, further preferably all, scanning points t.sub.s(n, m) to a respective adjacent scanning point is preferably different.

Description

[0070] The following description describes further principles, aspects, and embodiments of the invention, also with reference to the accompanying figures. Wherein:

[0071] FIG. 1 shows a schematic representation of a radar system according to the invention;

[0072] FIG. 2 shows an exemplary signal sequence;

[0073] FIG. 3 shows another schematic representation of a radar system according to the invention;

[0074] FIG. 4 shows another representation of a radar system according to the invention; and

[0075] FIG. 5 shows another representation of a radar system according to the invention.

[0076] In the description below, like reference numerals are used for like parts and parts having the same effect.

[0077] Basic structure of the radar system:

[0078] The radar system suitable for the method according to the invention can correspond in large parts to the structure of a common chirp sequence FMCW radar system. FIG. 1 shows a block diagram of a possible arrangement (wherein: SGEN=signal generator; SE=control unit; ADC=analog-to-digital converter; MIX=mixer; FLT=filter; OBJ=object/object structure; DSPE=digital signal processing unit):

[0079] The system comprises a signal generator SGEN for generating frequency modulated signals or for generating a sequence of a number of N frequency modulated signals s.sub.Tx(n, t) with n=1, 2, . . . , N, an antenna ATX for transmitting the signals s.sub.Tx(n, t) and an antenna ARX for receiving the reflected signals s.sub.Rx(n,t), wherein it is known to a person skilled in the art that it is also possible to use or configure one or more or all antennas both for transmitting and for receiving. The system further comprises a mixer MIX for mixing down the signal s.sub.Rx(n,t) with the signal s.sub.Tx(n, t), wherein it is generally known to a person skilled in the art that this mixer can be designed as a quadrature mixer (also called IQ mixer) as shown in the figure, but also as a single-channel real-valued mixer. In addition, a filter, preferably a low-pass filter, is used to filter the N downmixed signals s.sub.mix(n, t), wherein the filtered signals are called baseband signals s.sub.b(n,t). The system further comprises an analog-to-digital converter ADC for digitizing the baseband signal s.sub.b(n,t), wherein the N digital baseband signals are called s.sub.b(n,m), a digital signal processing unit DSPE, e.g. a microcontroller and/or microprocessor, a digital signal processor DSP, or a field programmable gate array FPGA, for processing the digital baseband signals, a control unit SE with which, the parameters of the frequency-modulated signals, in particular their sweep rate, can be set, on the one hand, and the scanning frequencies f.sub.s(n) with which the respective baseband signal s.sub.b(n,t) is scanned can be set on the other hand.

[0080] According to the embodiment, means are provided with which the value of the scanning frequencies f.sub.s(n) is directly coupled to at least one modulation parameter, such as the modulation rate, or to the sweep duration, or to the sweep bandwidth, of the N frequency-modulated signals, such that in particular the N frequency-modulated signals are scanned in a sequence with different scanning frequencies f.sub.s(n).

[0081] The radar system thus comprises an arrangement for generating and emitting a transmit radar signal sequence in the direction of an object scene and an arrangement for receiving and detecting the radar signal sequence reflected on the object structure, wherein said transmit radar signal sequence comprises a number of N frequency-modulated, preferably linear frequency-modulated, radar signals s.sub.Tx(n,t) (with n=1 N), wherein each of the N radar signals is characterized by a sweep start time t.sub.a(n) and start frequency f.sub.ca(n) and a sweep end time t.sub.a(n) and end frequency f.sub.ce(n) and a sweep rate μ(n) and a sweep duration T(n) (each with n=1 . . . N).

[0082] Wherein the sweep rate is defined as:

[00002]$\mu \left(n\right)=\frac{2.Math.\pi .Math.\left(f_{ce}\left(n\right)-f_{\mathrm{ca}}\left(n\right)\right)}{T\left(n\right)}\mathrm{with}T\left(n\right)=T_e\left(n\right)-T_a\left(n\right)$

[0083] and the sweep bandwidth is defined as: B(n)=|f.sub.ce(n)−f.sub.ca(n)|.

[0084] The N sweep rate values μ(n) of the frequency-modulated radar signals are not all equal with respect to their magnitudes and preferably all different. The N sweep rate values μ(n) are preferably selected from the two value intervals from μ.sub.min to μ.sub.max and from −μ.sub.max to −μ.sub.min, wherein μ.sub.max defines the amount of a maximum selected sweep rate and min defines the amount of a minimum selected sweep rate.

[0085] The N sweep rate values μ(n) can be selected in such a way that different transmit radar signal sequences, each with a different set of N values of μ(n) are formed and these different signal sequences have good orthogonality properties.

[0086] FIG. 2 shows an exemplary sequence.

[0087] The values of B(n) or the range of values in which all f.sub.ce(n) and all f.sub.ca(n) are preferably the same for all sweeps, as also shown in FIG. 2. However, they can also be selected differently for each sweep.

[0088] An advantageous selection of N sweep rate values μ(n), is to select the values of μ(n) as random values or as pseudo-random values from the aforementioned intervals, since this results in radar signal sequences with good orthogonality properties. In particular, two radar signal sequences have good correlation properties or good orthogonality properties if the N sweep rate values μ(n) in the two sequences differ as much as possible, e.g., if the sweep rate values are selected randomly distributed in the above intervals.

[0089] If two or more such signal sequences are mixed together (which can happen naturally in the present radar system in the receiver when several signal sequences are received simultaneously), the mixing products (i.e., the interference products) are statistically distributed almost uniformly over a broadband frequency range that may be determined by the selected sweep bandwidth.

[0090] A filter FLT preferably allows the desired baseband signal and also only that part of the interfering mixing products to pass that fall exactly within the frequency band or baseband defined by the filter FLT. However, since most of the interfering mixed products are outside the filter bandwidth, the interfering components can be significantly suppressed. If the design of two or more radar signals succeeds in such a way that, when the signals are correlated, their correlation products are statistically distributed over the maximum available bandwidth, the signals are optimal or at least improved with respect to their orthogonality properties or correlation properties.

[0091] The above signal sequences can be processed comparatively easily using the following steps.

[0092] Preferably (as above), the radar system comprises an arrangement for detecting a reflected radar signal sequence, with an antenna with which the reflected radar signal sequence is received and thus a receive radar signal sequence is formed, with a mixer MIX with which the receive radar signal sequence is mixed with the transmit radar signal sequence and thus N baseband signals s.sub.b(n,t) are formed, and these N baseband signals are scanned with an analog-to-digital converter (ADC) and then further processed as a digital signal s.sub.b(n, m).

[0093] As already partly explained before, N baseband signals s.sub.b(n, t) of a radar signal sequence from the analog-to-digital converter (ADC) with N different scanning rates f.sub.s(n) and the N values of the scanning rates f.sub.s(n) are selected to be proportional to the magnitude of the respective sweep rate μ(n).

[0094] Preferably, the scanning times t.sub.s(n,m) are chosen in such a way that the scanning points s.sub.b(n,m) with an identical index m have an identical carrier frequency value f.sub.c(m) (see FIG. 2).

[0095] By varying the scanning rates, it can be made possible that all (or at least a subset of the) scanned N baseband signals s.sub.b(n, m) have (exactly) the same number of scanning points in both dimensions despite their different time durations (which are determined by the respective sweep duration).

[0096] The N individual signals, formed from the scanning points sb.sub.n(m)=[s.sub.b(n,1), s.sub.b(n,2), . . . , s.sub.b(n,M),], correspond to the signals of common FMCW radars and could also be processed by common Fourier transform or other spectral analysis techniques to determine the distance and speed to a target. However, the usual two-dimensional processing of the baseband signal of a chirp sequence FMCW radar is not necessarily directly applicable because the scanning is not equidistant in the Doppler dimension.

[0097] The M individual signals in Doppler dimension direction are calculated from the sample points sb.sub.m(n)=[s.sub.b(1,m), s.sub.b(2,m), . . . , s.sub.b(N,m)] and, as can be seen in FIG. 2, each of these M signals is scanned non-equidistantly with a different scanning sequence.

[0098] This property, which at first appears to be a disadvantage, however, has an advantage with regard to the uniqueness range of the Doppler or speed measurement, as will be explained below. The following is a two-dimensional processing of the digital baseband signal according to the invention s.sub.b(n, m), which is particularly advantageous.

[0099] From the sample points s.sub.b(n, m), a number of M Doppler signals sb.sub.m(n)=[s.sub.b(1,m), s.sub.b(2,m), . . . , s.sub.b(N,m)] with N sample points each is formed, which is to be designated sb.sub.m(n) wherein the distance between adjacent scanning points is [t.sub.s(1,m), t.sub.s(2,m), . . . , t.sub.s(N,m)] is not equal and the scanning sequence [t.sub.s(1, m), t.sub.s(2,m), . . . , t.sub.s(N,m)] is therefore different for all M Doppler signals and the respective scanning sequence results from the choice of the sweep start times t.sub.a(n), the sweep start frequency f.sub.ca(n) the sweep end times t.sub.a(n) and the sweep end frequency f.sub.ce(n) (cf. FIG. 2);

[0100] M spectra Sb.sub.m(k) with K spectral values (k=1 to K) are first calculated from the M Doppler signals sb.sub.m(n), wherein the spectrum calculation is carried out using a method in which non-equidistantly scanned input values are permissible.

[0101] Advantageous methods that allow spectrum calculation with non-equidistantly scanned input values can be found, for example, in: [0102] Tarczynski, A., & Allay, N. (2004). Spectral analysis of randomly sampled signals: suppression of aliasing and sampler jitter. IEEE Transactions on Signal Processing, 52(12), pp. 3324-3334. [0103] Tropp, J. A., Laska, J. N., Duarte, M. F., Romberg, J. K, & Baraniuk, R. G. (2010). Beyond Nyquist: Efficient sampling of sparse bandlimited signals. IEEE Transactions on Information Theory, 56(1), pp. 520-544. [0104] Zandieh, A., Zareian, A., Azghani, M., & Marvasti, F. (2019). Reconstruction of sub-Nyquist random sampling for sparse and multi-band signals. arXiv preprint arXiv: 1411.6587.

[0105] Preferably all M spectra Sb.sub.m(k) on the same K discrete frequency base points k(k=1 to K) are calculated and in a second step, a total of K signals Sb.sub.k(m) is calculated from these M spectra Sb.sub.m(k) from all scanning points with the identical index k, and these K signals Sb.sub.k(m) (with m=1, 2, . . . , M) are subjected to another spectral analysis and thus K spectra Sb.sub.k(j) (with j=1, 2, . . . , J) or a matrix with the values Sb(k, j) is formed.

[0106] The other spectral analysis can be a fast Fourier transformation (FFT).

[0107] A distance Doppler diagram can be formed from the matrix with the values Sb(k, j), with which diagram target objects are detected and a distance or speed is assigned to them.

[0108] It is known from the theory of spectral analysis with non-equidistantly scanned signals that the spectra of such signals, unlike those of equidistantly scanned signals, do not repeat periodically and, at favorable scanning frequencies, frequencies significantly higher than half the average scanning frequency can be estimated correctly and unambiguously. It is also known that scanning frequencies with statistically or randomly selected scanning times, as they can be present here in the Doppler dimension, are advantageously suited for unambiguous frequency estimation.

[0109] For this reason, in particular, a correct and unambiguous estimate of the speed of a radar target is made possible over a much wider range of uniqueness than would be possible with prior art chirp sequence FMCW radars with similar sweep rates and similar sequence durations. This is a great advantage especially for the use of such radars in the automotive, railroad or aviation sector or also for many other applications where objects with a wide range of possible speeds can be located in the detection range of a radar.

[0110] Methods in which the bandwidth is varied (along with the scanning rate) can give the same effect (although not necessarily with the advantages of extended Doppler range, but possibly with the advantages of a larger unambiguous range of distances; the two approaches can also be combined if necessary)

[0111] Since, as explained above, the signal sequences in particular also have good orthogonality properties or good correlation properties, they can also be used advantageously in MIMO radars to enable simultaneous transmission on multiple transmit channels.

[0112] FIGS. 3, 4, and 5 illustrate one possible embodiment of such a radar.

[0113] The system comprises a number of Q signal generators (index q=1 . . . Q) for generating the signal sequences according to the invention and Q antennas (index q=1 . . . Q) for transmitting these signal sequences. The Q signal sequences are now all to be transmitted simultaneously. It is also conceivable to send only subgroups, i.e. a limited number, e.g. 2 or 3, of the Q signal sequences simultaneously, which could reduce the effort in the receiver, but is not crucial here. In accordance with embodiments, the transmitter circuit behind each of the Q signal generators is designed to distribute the transmit signal to a number of R paths using a signal divider (1 on R) adjacent to the path on which it is transmitted.

[0114] The embodiment further comprises a number of R receiving antennas (index r=1 . . . R) for receiving the radar signals. According to the embodiment, the receiver circuit behind each of the R receiving antennas is designed in such a way that the receive signal is distributed to a number of Q paths using a signal divider (1 on Q).

[0115] The receiver circuit behind each of the R receiving antennas can be designed in such a way that the receive signal of a receiving antenna is fed to a number of Q mixers (or to as many mixers as signals were transmitted simultaneously). As the signal used for mixing down, the Q mixers are each fed with one of the Q signal sequences. The mixed signals are filtered with a filter FLT before further processing of the signals is performed.

[0116] FIG. 4 shows a sketch of a possible embodiment of a section of the complete circuit with only one of the transmit channels (here the transmit channel with the index q). The circuits for all other transmit channels can be designed in the same way.

[0117] FIG. 5 shows a sketch of a possible embodiment of a section of the overall circuit with only one of the receive channels (here the receive channel with the index r). The circuits for all other receive channels can be designed in the same way.

[0118] Finally, the receive signal of a receive antenna (index r) is mixed down with all Q signal sequences, filtered, scanned with an ADC, such that for each of the Q transmit channels R digital baseband signals s.sub.b(n,m,q,r=1 . . . R) are formed for each of the R receive channels Q (cf. FIG. 4), or digital baseband signals s.sub.b(n,m, q=1 . . . Q,r) are formed (cf. FIG. 5), i.e., ultimately a total of R times Q digital baseband signals s.sub.b(n, m, q=1 . . . Q, r=1 . . . R) are formed.

[0119] Due to the good orthogonality properties or the good correlation properties of the signal sequences according to the invention, only the signal components of one transmit signal at a time, and always of the one with which downmixing was performed, are strongly pronounced, and the others are strongly suppressed, in the Q distance Doppler diagrams which are formed from the Q mixed signals of a receive channel (i.e. those from one of the receive antennas). Thus, despite simultaneous transmission with multiple transmit signals, the resulting receive signals can be processed separately for each of the R by Q transmit-receive antenna combinations.

[0120] It is also possible to scan the R receive signals directly and perform the mixing processes in digitized form in a computing unit.

[0121] It should be noted at this point that all of the parts described above, considered on their own and in any combination, in particular the details shown in the drawings, are claimed as possibly also independent inventive ideas, in particular as respective independent further developments of the inventions shown in the claims. Amendments to this are possible.