WO2020120334A1

WO2020120334A1 - Method for a mimo radar for unambiguously determining targets, and a radar

Info

Publication number: WO2020120334A1
Application number: PCT/EP2019/084072
Authority: WO
Inventors: Benjamin Sick; Stefan Zechner; Florian Engels; Adam HEENAN; Ernest CASABAN
Original assignee: Zf Friedrichshafen Ag
Priority date: 2018-12-10
Filing date: 2019-12-06
Publication date: 2020-06-18
Also published as: DE102018221282A1

Abstract

A method for a MIMO radar for unambiguously determining targets, wherein the radar comprises a plurality of transmitting antennas and a plurality of receiving antennas, • wherein each transmitting antenna emits a plurality of temporally successive frequency ramps (54, 56), • wherein the start phases (φ1, φ2) of successive frequency ramps of the transmitting antennas are varied in accordance with a periodic coding, • wherein the receiving antennas detect the frequency ramps (54, 56) reflected at a target as reception signals, • wherein the reception signals are processed in order to provide a spectrum (40), • wherein the energy of a detected target within the spectrum (40) is distributed among a plurality of positions (42, 44, 46, 48) on account of the coding, • wherein the coding is chosen such that the energy distribution for which each position (42, 44, 46, 48) has an energy value is distributed unambiguously over the positions (42, 44, 46, 48), • wherein a target is determined from the spectrum (40) on the basis of the unambiguous energy distribution. A radar that uses this method is additionally described.

Description

Procedure for a MIMO radar for the clear determination of targets as well as a radar

The invention relates to a method for a MIMO radar for the unambiguous determination of targets and a radar which is designed for this method.

Radars which use MIMO methods are known in the prior art. These include a plurality of transmitting antennas and a plurality of receiving antennas. The successive frequency ramps of the transmitting antenna are encoded by specifically varying the start phase. The coding is often selected periodically here. A radar signal reflected from an object is thereby distributed over several positions in the determined spectrum. Particularly when using multiple transmit antennas, a plurality of energy peaks are generated in the spectrum, all of which belong to the same target of the object. It can occur in certain cases that it is no longer possible to clearly differentiate between the energy components of the different transmission antennas. As a result, the amplitude and the angle of the target may be incorrectly determined.

It is the task to provide a method for a MIMO radar and a corresponding radar, which enables a clear determination of targets.

This object is achieved by a method according to the features of claim 1. Advantageous embodiments are explained in the dependent claims.

The method is suitable for a MIMO radar (multiple input multiple output radar). Such a radar comprises several transmitting antennas and several receiving antennas. In addition, the radar preferably works as an FMCW radar (Frequency Modulated Continuous Wave radar).

According to the method, each transmitting antenna transmits a number of frequency ramps which follow one another in time. These frequency ramps are also called chirps or ramps. The successive frequency ramps of each transmitting antenna are varied according to a periodic coding. This means that the start phase of each frequency ramp sent out is set specifically for each frequency ramp. The specifically set start phase, which varies for successive frequency ramps, is also referred to as the coding sequence for a single transmission antenna. The term coding thus encompasses the coding sequences of all transmission antennas. The individual NEN antennas preferably have different coding sequences with respect to coding. The coding sequences are periodically applied to the frequency ramps of the individual transmission antennas. The coding sequence follows periodically with a periodicity with the value N. N indicates the minimum number of start phases for the successive frequency ramps, the coding sequence necessary to provide for a transmitting antenna. In other words, N describes the number of frequency ramps before the coding sequence is repeated. If at least 4 defined successive start phases for the frequency ramps are necessary for a coding sequence, then this has a periodicity of 4.

The phase is varied in binary form, that is to say at 0 ° or 180 °. This is also known as BPSK (Binary Phase Shift Keying). Alternatively, the variation of the phase jumps can also be chosen freely, which is also called PSK (phase shift keying).

Different transmission antennas work with different coding sequences during transmission, for example the coding sequence [0000] for a first transmission antenna, which can also be reduced to [0], and the code [0011] for the second transmission antenna. The coding sequences of the two transmit antennas chosen as examples have a periodicity of 1 and 4.

Radar signals emitted, in particular the frequency ramps coded by each transmitting antenna, can be reflected on an object that is within the field of view of the radar. The reflected radar signals are then detected by the receiving antennas as input signals. An object corresponds to an object that has one or more scattering centers. The scattering centers can be detected by the radar as targets, in particular via a or several dimensions of relative distance, relative speed, azimuth angle and / or elevation angle is resolved. This is done by evaluating the

Receive signals. Detected scattering centers or reflection centers are also referred to as targets.

A spectrum is determined from the received signals by evaluation. The determination of such a spectrum is well known in the prior art. This is usually done via one or more Fourier transformations, depending on which dimensions are used to resolve. The spectrum can accordingly be multidimensional. The spectrum is 4 dimensional for resolution via distance, speed, azimuth angle and elevation angle. Targets appear as energy peaks within the spectrum.

The energy of a detected target is distributed over several different positions within the spectrum due to the coding. This means that several energy peaks occur within a multidimensional spectrum, all of which belong to the same goal. The distribution depends on the number of transmit antennas and the coding.

Two transmission antennas with the same transmission power are selected as examples. Accordingly, in the spectrum, a total energy value of 2 occurs for a target. This total energy value is distributed across the various positions. In example, the first antenna is coded so that it maps to a single position. This position then includes the energy value 1. The second antenna can have a coding that maps to two positions, so that an energy value 0.5 falls in two positions. The coding sequence, in particular its periodicity N, determines the number of positions to which the associated energy value is distributed in the spectrum.

The distribution of the total energy value over the different positions depends on the transmission energy of the transmission antennas and the chosen coding. In particular, the transmit antennas can also have different transmit powers, so that the total energy value is no longer an integer. The example corresponds to Total energy value when using 4 transmitting antennas, all of which transmit with the same power, a value of 4.

The number of positions depends on the coding sequence. With a periodicity of 2 the total energy value is divided into 2 different positions, with a periodicity from 4 to 4 different positions, etc. An energy value does not have to occur at every position.

The coding is chosen such that the energy distribution, in which each position has an energy value, is uniquely distributed over the positions. In other words, a defined energy value occurs for a target at the associated positions. The energy values have a certain relationship to one another, so that conclusions can be drawn unambiguously and these can also be clearly assigned to the respective transmission antennas. The coding is therefore only clear if it is achieved for a real measuring process that at least the ratio between at least two energy values in the spectrum can be clearly assigned to the expected energy distribution.

This is especially not the case if the total energy value is distributed across all positions and the expected energy values for all positions differ only slightly. A slight distinction is made when the energy values are so close apart that, under certain circumstances, e.g. caused by noise, an actually lower energy value can be measured larger at one position than an actually larger energy value at another position.

Such an unambiguous distribution thus arises precisely when, in addition to high energy values, lower energy values also occur at the positions. In particular, an energy value at a position can also be zero. Accordingly, a missing position can also be configured as information. This makes it possible to clearly verify whether an energy value in the spectrum is a real target. In addition, it can be clearly assigned from which transmission antenna the respective energy components originate. Thereby goals can be clearly defined. There is a gap between low and high energy values if, in a real measurement, the higher energy value remains identifiable larger than the low energy value.

Accordingly, a target from the spectrum is determined based on the clear energy distribution.

This can be done by analyzing the spectrum directly for the corresponding clear energy distribution. Alternatively, an energy value can also be determined at a specific position and then verified whether the corresponding energy values are available at the associated further positions of the energy distribution.

When determining the target, corresponding ranges for the energy value can be defined for each position of the distribution, within which the actually measured energy values may lie. This allows external circumstances, such as noise, to be taken into account. If the energy value of all associated positions for a distribution lies within the associated areas, the target is accepted as correct. Otherwise the target is discarded.

For example, a coding is assumed in which a total energy value is distributed from two to four positions in the spectrum. A transmitting antenna is selected in such a way that it maps the energy value 1 to the first position in the spectrum. The second antenna maps to the second and fourth positions with an energy value of 0.5 each. At the third position in the spectrum, the energy value is 0. The energy distribution is [1, 0.5, 0, 0.5].

When evaluating the targets, a maximum is now detected which corresponds to the first position. For verification, it is then determined whether there is a relative energy value of 0.5 at the second position, an energy value of 0 at the third position and an energy value of 0.5 at the fourth position. If this is the case, the target is clearly identified and the solid angle can be determined correctly. Treads deviating at the associated other positions Energy values, the energy value in the spectrum is not a real goal but an incorrect assignment of the energy values.

The spectrum is analyzed with particular advantage directly after the expected energy distribution for the goals.

Advantageous embodiments of the invention are explained below.

It is proposed that at least one energy value of the energy distribution is selected or is zero in comparison with the energy values of the other positions with comparatively low energy.

The distribution of the total energy value over the positions in the spectrum can be clearly determined, in particular by means of comparatively low energy values. Up to now, a single maximum energy value has been provided in the case of encodings which have been customary to date, the energy values of the other positions also having high values. In most cases, these differ only by a few dB, which is why, under certain circumstances, no clear evaluation of the target is possible.

It is particularly advantageous if at least one energy value is at least 50%, 70% or 90% smaller than the next largest energy value or the highest energy value.

Such an energy value difference can be clearly determined in the spectrum. With differences of 1 - 2 dB, as is common with current MIMO radars, a clear distinction is not always possible due to measurement effects.

Favorably, at least one energy value is less than or equal to 0.4, 0.3, 0.2, 0.1, 0.05 or 0.

The energy value relates accordingly to the energy value of a transmitting antenna ne, which maps the spectrum with a value of 1. In particular, low energy values can greatly improve the uniqueness of the evaluation. When using With 4 antennas with a total energy value of 4, the target can be optimally verified with such a low energy value at one position.

In a further embodiment variant, at least one transmission antenna has a coding sequence of the frequency ramps with a periodicity of N, the periodicity being greater than the number of transmission antennas. A plurality of transmitting antennas advantageously have such a coding sequence.

The total energy value of all transmit antennas is selected such that an at least low energy value is provided for at least one position with respect to the distribution over the individual positions, which depends on the coding.

If the number of transmit antennas is high compared to the selected periodicity, there are not enough positions in the spectrum to distribute the total energy value in such a way that low energy values are possible. Accordingly, in such an embodiment, all positions have high energy values.

If the periodicity is chosen to be greater than the number of transmit antennas, the total energy value can be distributed over correspondingly more positions, so that one or some of the positions also have low energy values, which enable a clear evaluation. The periodicity is chosen larger than the number of transmit antennas for a plurality of transmit antennas. In particular, at least one third, half or three quarters of the transmission antennas have a periodicity that is greater than the number of transmission antennas.

The periodicity for one or more transmit antennas is at least twice as large as the number of transmit antennas.

In a further embodiment variant, the transmit antennas transmit with different transmit power. This means a different transmission power, in particular an average transmission power over all frequency ramps of a measurement cycle. The transmission power for the successive frequency ramps can be identical or can vary. If you combine the defined choice of the mean transmission power for the individual transmission antennas with a favorable coding, the distribution of the total energy value over the different positions in the spectrum can be selected in such a way that the clear evaluation of the targets is even more robust. Accordingly, each transmitting antenna can transmit the radar radiation with a different transmission power.

It is further proposed that, in addition to the start phase, the transmission power of the transmission antenna is also varied on successive frequency ramps.

In this way, the distribution of the energy values across the various positions can be adjusted even further. In particular, the robustness of the clear evaluation of the goals can be further improved.

The object is also achieved by a radar according to claim 9.

The radar is designed to carry out the procedure described above and also in the following. In particular, this can be carried out in accordance with at least one of the explanations given above and also below.

The method and the radar are illustrated below using several figures. Show it:

1 spectrum for coding 0001;

2 spectrum for coding 0011;

Fig. 3 frequency ramps of the transmit antennas;

Fig. 4 flow chart for the method. The method is described using a MIMO radar that works according to the FMCW method. The spectra shown represent the results of a simulation for a single static target within the field of view of the radar. In addition, for the sake of a descriptive description, the radar only comprises two transmitting antennas, each of which emits several frequency ramps in succession at uniform time intervals within a measurement cycle. The outside of the frequency ramps corresponds to a first step 100 of the method, the individual steps of the method being shown in FIG. 4. The radar beams emitted in the form of frequency ramps can be reflected on an object, in particular on a scattering center. Here, for example, a single static target is selected which has a relative speed of zero compared to the radar. According to step 110, the reflected radar beams are detected by a plurality of reception antennas as reception signals.

The received signals are processed in the usual way in order to provide a spectrum according to step 120. The spectrum can basically resolve over one or more dimensions, such as distance, speed, azimuth angle or elevation angle. 1 and 2 simulated one-dimensional spectra are shown, which serve to further explain the method.

The spectra represent the result for a single target, which has a relative speed of 0 km / h compared to the radar. In this case, only a one-dimensional resolution is chosen in favor of a simple and descriptive explanation of the method. In the real application, a plurality of targets are detected within a field of view of the radar and the resolution takes place over several dimensions, in particular distance, speed, azimuth angle and elevation angle.

Compared to the respective X axes 10 of FIGS. 1 and 2, a normalized frequency is plotted, which corresponds to a speed of the target or can be converted into a relative speed. The energy is plotted against the Y axis 12. 3, some frequency ramps 54, 56 are shown, which are emitted by the two transmission antennas Tx1 and 1x2. The time is plotted against the X axis 50 and the frequency is plotted against the Y axis 52. For example, in one measurement cycle, 256 transmit frequencies are transmitted from each transmit antenna to one another. A frequency ramp 54, 56 corresponds to a radar signal that begins at a start frequency and ends with an end frequency that differs from the start frequency. As shown in FIG. 3, the frequency of the frequency ramps 54, 56 increases linearly from a start frequency to the end frequency. In this case, the frequency ramps 54, 56 are emitted for each transmitting antenna at regular intervals.

The frequency ramps are transmitted under coding. In the coding, the start phases f _{1 of} the frequency ramps 54 of the transmission antenna Tx1 and the start phases f _{2 of} the frequency ramps 56 of the transmission antenna Tx2 are varied. The start phase is the phase position which the frequency ramp has at the start of the transmission, that is to say at the start frequency. Basically, more than two transmit antennas can be formed, each transmit antenna emitting its frequency ramp with its own coding sequence.

The spectrum of FIG. 1, which is represented by line 30, results with reference to a target which has a relative speed of 0 km / h with respect to the radar, the transmitting antenna Tx1 having a coding sequence of [0000]. This means that the start phase f ₁ is identical for all successive frequency frames. In particular, the start phase is 0 °.

The coding sequence of the transmitting antenna Tx2 is [0001]. This corresponds to a periodicity of 4, since four frequency ramps are necessary for the passage of the coding sequence. The coding sequence is applied periodically, ie repetitively, to the frequency ramps of the transmission antenna Tx2. The start phase <p ₂ corresponds to a start phase of 0 ° for successive frequency ramps with a value of 0 and a start phase of 180 ° for value 1. This is also known as binary phase shift keying, BPSK, in which the start phase varies between two start phases offset by 180 °. Basically, phase shift keying, PSK, can be used, in which the start phase can be set to any start phase between 0 ° and 360 ° and varied accordingly.

The transmitting antenna Tx1 generates an energy value at a position 32 within the spectrum, which represents an absolute maximum for the spectrum. In addition, the transmit antenna Tx2 generates corresponding and essentially identical energy values at positions 34, 36 and 38. Overall, the total energy emitted by the transmitting antennas is divided into positions 32, 34, 36 and 38. The spectrum is now evaluated, the energy values of the individual positions identifying the transmitting components of transmitting antenna Tx1 and transmitting antenna Tx2.

Knowing which energy value comes from which antenna is necessary because if the peaks are mixed up or swapped, a subsequent angle determination would be incorrect. In the case of a real measurement, it may happen that the energy value at position 34, 36 or 38 is not determined as belonging to the transmitting antenna Tx1, but rather the energy value 32. The subsequent angle determination would be incorrect. Such incorrect evaluations can occur if, due to external circumstances, for example due to reduced spectral calculations, noise effects or a plurality of targets in the vicinity of the position under consideration, the energy value of the transmitting antenna 1 is lower and an energy value of another position is higher. This coding is therefore error-prone. The spectrum of Fig. 1 describes the prior art approach.

This problem is solved by a clear coding, which provides a spectrum on the basis of which goals can be clearly evaluated. In particular, the proportions of the transmitting antennas can be clearly assigned.

Such a spectrum is shown on the line 40 in FIG. 2. The coding sequence of the transmitting antenna Tx1 is identical to that in FIG. 1. The coding sequence of the transmitting antenna Tx2 is [0011]. As a result, the total energy value of the transmit antennas is distributed to positions 42, 44, 46 and 48, the largest energy value at position 44 and two essentially identical energy values at positions 44 and 48. At position 46 at a normalized frequency of 0.5, no energy components were shown. Due to the chosen coding, the distribution of the total energy value in the spectrum is known. The associated energy distribution is [1, 0.5, 0, 0.5]. This makes it possible to verify whether the energy values at the respective positions have been correctly determined for the further evaluation of the azimuth and / or elevation angle. With regard to an energy value that is generated in the spectrum by the transmitting antenna Tx1, only a low or no energy value may occur at a position that is further by a normalized frequency of 0.5. In principle, the sequence of energy values for the individual energy peaks must correspond to the distribution and only then is it a real and correctly determined target. With a target determined in this way, there is also the certainty that subsequent determination of the azimuth angle and / or elevation angle is correct.

In a real measurement, it could be the case due to external circumstances that the energy value at position 44 is greater than the energy value at position 42. Since position 48, which is 0.5 normal frequency further, also has a high energy value , it can be excluded that this is a correctly determined target. Accordingly, there is no subsequent incorrect determination of targets. In this exemplary embodiment, only two transmit antennas are described, a radar in real measurement mode supposedly having more transmit antennas. Due to the larger number of transmit antennas and the associated total energy value, the energy values at the individual positions are much closer together, so that erroneous determination of the targets occurs more often.

With a suitable choice of energy ranges for the individual positions of the energy distribution, the target can nevertheless be correctly recognized and its position and relative speed in the field of view of the radar can be correctly determined. The target is accepted as correct if the determined energy values at the positions lie within the energy range. In principle, the energy areas can be fixed static areas. The energy ranges can also be fundamentally can also be determined separately for each distribution by using an algorithm

Checks energy distribution for plausibility.

By choosing a clear coding and its corresponding coding sequence, the distribution of the total energy values of the transmitting antennas in the spectrum can be set in such a way that the goals are clearly and correctly evaluated. In this way, erroneous detection of targets can be reduced by more than 10% compared to the method of the prior art.

For example, when using 4 transmit antennas that provide a total energy value of 4 within a spectrum for a target, a coding can be used that has an energy distribution of [1; 0.5; 0.5; 0.5; 0; 0.5; 0.5; 0.5] ready. The first position has an energy value of 1, the second position has an energy value of 0.5, the third position has an energy value of 0.5, the fourth position has an energy value of 0.5, the fifth position has an energy value of 0, etc. As soon as a potential energy value for a target has been determined, the corresponding positions and their energy values can be used to verify whether the target has been correctly determined. Alternatively, the target can be determined directly via the energy distribution of the energy values for the corresponding positions. For the distribution of the total energy value over 8 positions, for example, a periodicity of 8 is used for a coding sequence. For example, if a target with an energy distribution of [1; 0.43; 0.59; 0.55; 0.08; 0.52; 0.48; 0.5] so the ses could be accepted as correct.

The number of positions and / or the periodicity of at least one or more transmit antennas is advantageously greater than the number of transmit antennas. Due to the larger number of positions compared to the number of transmit antennas, a distribution of the total energy value can take place, in which comparatively low energy values also occur. It is therefore generally advantageous if the number of positions or the periodicity is at least twice as large as the number of transmit antennas. With 16 transmit antennas, a coding with 32 positions is advantageously chosen. The distribution can be such that the energy value for one or more positions is zero or has only comparatively small values. With regard to a transmission antenna with a transmission energy of 1, this can be selected, for example, as an energy value of 0.4, 0.3, 0.2, 0.1, 0.05 or 0. The individual transmission antennas can also be operated with different transmission power in order to provide correspondingly clear distributions of the total energy value in the spectrum.

A target is accordingly determined on the basis of the clear energy distribution according to step 130. This ensures that the determined target is a real and correctly detected target.

Reference sign

10 X-axis (frequency)

12 Y axis (energy)

30 line (spectrum)

32 position

34 position

36 position

38 position

50 X-axis (time)

52 Y-axis (frequency)

54 frequency ramp (Tx1)

56 frequency ramp (Tx2)

100 paces

110 step

120 step

130 step

f _ί start phase

f ₂ start phase

Tx1 transmission antenna

Tx2 transmission antenna

Claims

1. A method for a MIMO radar for the unambiguous determination of targets, the radar having a plurality of transmitting antennas and a plurality of receiving antennas,

• wherein each transmitting antenna emits a number of frequency ramps (54, 56) that follow one another in time,

The starting phases (f ₁ , f ₂ ) of successive frequency ramps of the transmission antennas are varied according to a periodic coding,

Wherein the receiving antennas detect the frequency ramps (54, 56) reflected on a target as received signals,

Wherein the received signals are processed to provide a spectrum (40),

Wherein the energy of a detected target within the spectrum (40) is distributed over several positions (42, 44, 46, 48) due to the coding,

Wherein the coding is selected such that the energy distribution, in which each position (42, 44, 46, 48) has an energy value, is uniquely distributed over the positions (42, 44, 46, 48),

• a target from the spectrum (40) is determined on the basis of the clear energy distribution.

2. The method according to claim 1, characterized in that at least one energy value (46) of the energy distribution is selected in comparison with the energy values (42, 44, 48) of the other positions with comparatively low energy or is zero.

3. The method according to claim 1 or 2, characterized in that at least one energy value is at least 50% smaller than the next largest energy value or the highest energy value.

4. The method according to any one of claims 1 to 3, characterized in that has at least one energy value at a position less than or equal to 0.4, 0.3, 0.2, 0.1, 0.05 or 0.

5. The method according to any one of claims 1 to 4, characterized in that at least one transmitting antenna has a coding sequence with a periodicity of N, wherein the periodicity N is greater than the number of transmitting antennas.

6. The method according to claim 5, characterized in that the periodicity N for one or more transmit antennas is at least twice as large as the number of transmit antennas.

7. The method according to any one of claims 1 to 6, characterized in that the transmission antennas transmit with different transmission power.

8. The method according to any one of claims 1 to 7, characterized in that during the coding in addition to the start phase of successive frequency ramps (54, 56), the transmission power of the transmission antenna is also varied.

9. radar designed to carry out the method according to one of claims 1 to 8.