Method for self-calibration of a radar system

Abstract

The invention relates to a method for the self-calibration of a radar system (RS) with at least two antenna groups to which at least one transmission channel and one receiving channel are assigned. To start with, a plurality of targets are measured (10) by the radar system (RS). Then, for each antenna group, the distance and Doppler information (11) is processed, and the targets are detected (12), in order to obtain reflex lists (L) with complex amplitudes for each target. Subsequently, the amplitude differences for the respective channels are compensated (20). By means of a two-dimensional linear regression (21), a regression plane, for which the average quadratic distance of the phase measurement values of the channels of the antenna groups is minimal, is then estimated and the difference between the measured phase value and the regression plane is calculated for each channel in order to obtain an intra group phase correction value (Kintra). In addition, the distance between two regression planes of different antenna groups is calculated (25) with modulo 2π in order to obtain an inter group phase correction value (Kinter). Finally, the control vector of each channel is compensated (26) with the intra group phase correction value (Kintra) and the inter group phase correction value (Kinter).

Description

[0001] Description

[0002] title

[0003] Procedure for self-calibration of a radar system

[0004] The present invention relates to a method for self-calibration of a radar system having at least two antenna groups. At least one transmit channel and one receive channel are assigned to each antenna group.

[0005] State of the art

[0006] Radar systems are used to measure the distance, relative speed, and azimuth and elevation angles of objects. Antenna groups, also known as antenna arrays, are typically used for angle estimation. These can be used for both transmitting and receiving. Digital beamforming is used to detect the angle of incidence of a reflected plane wave, such as that generated by targets in the far field. Beamforming involves evaluating the phase differences of the reflected wave across multiple receive channels. The received signals from the antenna groups can be transformed to appear as if they had been measured by virtual receivers. The evaluation is traditionally performed using pre-stored control vectors, which are applied in various methods.For example, in the Bartlett beamformer, the expected phase differences for different angles of incidence are stored in the control vectors, which are then correlated. Alternatively, model-based estimates can be performed, which are also based on the control vectors.

[0007] Traditionally, the control vectors are measured over the angle (one or more angle sections) for each individual sensor during a one-time end-of-line calibration and then stored in non-volatile memory. This takes into account a phase error (phase offset) that can arise from various effects, such as manufacturing tolerances of antennas or feed lines, or interaction of the electromagnetic waves with a radome, a housing, or a circuit board.

[0008] The antenna arrays also allow the environment to be divided into angular cells and detected separately. This allows targets with identical range and radial velocity to be separated from each other based on their angular position. Each radar beam of the radar system represents a spatial filter.

[0009] For the specific functions of vehicle automation, better angle measurements are required in terms of both accuracy and selectivity. To achieve this, large-aperture antenna arrays are used. To control the corresponding transceivers, these are typically grouped and cascaded, creating a transceiver hierarchy. To perform a coherent measurement of the angle of incidence of the reflected wave, all transceivers are preferably connected to a reference oscillator. Radio-frequency lines are typically used for this purpose, which distribute the reference oscillator signal to the transceivers. However, the larger the aperture, the longer the radio-frequency lines.

[0010] Today, patch antennas are commonly used as radar antennas in the automotive sector. Patch antennas are mounted on the surface of a circuit board and connected via striplines. Consequently, the size of the circuit boards increases proportionally to the size of the aperture. With larger circuit boards, the likelihood of an inhomogeneous temperature distribution on the board, caused by internal and / or external factors, also increases. An example of an internal factor is the significant heating of various component groups. Examples of external factors include partial shading of the circuit board and heating due to neighboring components and / or air currents.

[0011] The inhomogeneous temperature distribution, component aging, and other factors affect the radar system's measurements by altering the pre-calibrated phase errors. If angle estimation continues to be performed using the initially stored control vectors, the performance of the angle estimation is reduced. In particular, the side lobes increase in the angular spectrum, while the main lobe is attenuated and broadened, and the position of the main lobe changes (beam-pointing error). As a result, the dynamic range, accuracy, and discriminatory power of the angle estimation are reduced.

[0012] A self-calibration of a radar system is known from the article by M. Harter et al., "Error analysis and self-calibration of a digital beamforming radar system," 2015 IEEE MTT-S International Conference on Microwaves for Intelligent Mobility (ICMIM), 2015, pp. 1-4. It assumes that the wavefronts of the reflected wave at the antennas are flat. This assumption is only sufficiently fulfilled if the target is in the far field. The boundary for the far field is typically defined using Formula 1 for the aperture length D and the wavelength A:

[0013] (Formula 1)

[0014] Assuming that the target is in the far field, a one-dimensional linear regression is performed using the received phases at the antenna elements of the receiving antenna arrays, thus estimating the most probable wavefront path. This is then used to calculate the phase errors to be corrected. The phase errors estimated from a target apply to the respective target angle without loss of generality. It is irrelevant whether the radar system has a radome or is mounted behind a bumper, for example.

[0015] A further prerequisite for self-calibration according to the above-mentioned article is that the antenna elements are arranged equidistantly along one dimension. This arrangement is as follows: For calibration of the azimuth control vectors, the antenna elements are arranged horizontally, and for calibration of the elevation control vectors, the antenna elements are arranged vertically.

[0016] In general, self-calibration can be used for various modulation methods. Typical transmission frequencies today are 24 GHz or 77 GHz, and the maximum usable bandwidths are below 4 GHz, especially around 0.5 GHz.

[0017] Today's vehicle-mounted radar systems typically use FMCW (frequency-modulated continuous wave) modulation with fast ramps (fast chirp modulation), in which several linear frequency ramps of the same gradient are transmitted one after the other. Mixing the instantaneous transmitted signal with the received signal produces a low-frequency signal whose frequency (called the beat frequency) is proportional to the distance. The system is generally designed so that the component of the beat frequency caused by the Doppler frequency is negligible. The distance information obtained from the beat frequency is largely unambiguous; a Doppler shift can subsequently be determined by observing the temporal evolution of the phase of the complex distance signal over the ramps. The distance and speed are determined independently of each other, usually using two-dimensional Fourier transformation.The angle estimation described above is downstream of the distance and speed estimation.

[0018] Disclosure of the invention

[0019] A method for self-calibration of a radar system with at least two antenna groups, each assigned at least one transmit channel and one receive channel, is proposed. The geometry of the antenna groups, i.e., the position of the transmit and receive antennas, is known. Furthermore, the transceiver hierarchy of the transceivers used to form the antenna groups, whose elements experience similar phase errors, is known. This data is part of the radar system's specification.

[0020] Initially, several targets are measured by the radar system. For this purpose, transmitting antenna groups emit electromagnetic waves in the direction of the targets. The electromagnetic waves reflected by the targets are then picked up by receiving antenna groups and evaluated using digital beamforming. The measurement is carried out with angular resolution, with the surroundings being divided into angular cells by the antenna groups. The range and Doppler information is processed for each antenna group (also known as range-Doppler processing). The processing is preferably carried out using a two-dimensional fast Fourier transform, but can also be performed using other known algorithms. The targets are then detected in the angular cells. For this purpose, the following steps are preferably carried out: The energy is determined for each angular cell.In addition, a threshold value is estimated for the estimated noise energy of such an angular cell. The determined energy is compared with the threshold value for the noise energy, thus discriminating the determined energy in each cell from the estimated noise energy. If several neighboring angular cells are above the threshold value for the estimated noise energy, the local maxima of the measured complex data are determined for each combination of transmitters and receivers. The above steps produce reflex lists in which the distance, relative velocity, and complex amplitudes measured on the respective virtual channels are stored for each target.

[0021] First, the amplitude differences are compensated for the respective channels. The following steps are preferably performed for this purpose. The average received power across all signal amplitudes is calculated, using Formula 2 in particular:

[0022] (Formula 2)

[0023] |x| represents the mean of the amplitudes and x mn the individual measured complex amplitudes. The summation is carried out over the transmitters m = 1...M and the receivers n = 1...N.

[0024] The deviation of the amplitude from the average received power is then calculated for each channel. Specifically, the deviation is calculated using a quotient according to Formula 3:

[0025] (Formula 3)

[0026] The deviation is then used to compensate for an amplitude error for each channel. This step allows for calibration of the received power.

[0027] Next, an intragroup calibration is performed for each antenna group. For this purpose, a two-dimensional linear regression is performed for each antenna group. The regression plane is estimated for which the mean square difference between the phase measurement values ​​of the channels of the antenna groups is minimized. The two-dimensional regression uses the geometric data of the antenna groups, which, as described above, are known as part of the radar system specification. For each channel, the difference between the measured phase value and the regression plane is calculated. This yields the intragroup phase correction value for the respective antenna group. This step is then repeated for all antenna groups so that intragroup phase correction values ​​are available for all (virtual) channels.

[0028] In the event that the phase error is not angle-dependent - for example, if the phase field is due to aging of the radar system - the measured values ​​of all targets for each channel can be averaged. This achieves an averaging gain. In the event that the phase error is angle-dependent - for example, because the radar system is located behind a bumper - each angle can be calibrated using its own target. Preferably, intermediate angles for which no separate target is available can be interpolated. In this case, the angles for which targets are present should be close enough to each other so that the sampling theorem is observed for the phase error. By running several measurement cycles in succession, targets can be detected in angle cells that were previously only interpolated or for which no phase error estimation was yet possible.

[0029] Next, an intergroup calibration is performed for at least two different antenna groups to determine the phase error between the different antenna groups. A regression plane was determined for each antenna group as described above. The distance between two regression planes of two different antenna groups is then calculated, applying modulo 2n. This yields the intergroup phase correction value for the two antenna groups. If the regression planes are compensated with the intergroup phase correction value, all regression planes of the antenna groups are aligned.

[0030] Finally, the control vectors of each channel are compensated with the intragroup phase correction value and the intergroup phase correction value. Angle estimation can then be performed using a known method.

[0031] Due to the two-dimensional regression, the method can generally be applied to all antenna array geometries, and these are not subject to any restrictions regarding either alignment or spacing. The antenna arrays do not necessarily have to be arranged horizontally or vertically, but can be freely positioned. Furthermore, the antenna elements do not have to be equidistant from each other, but can be spaced at any desired distance. Furthermore, the phase errors for each channel do not have to be uniformly distributed; a different average phase error is possible for each antenna array. The method can also be used for multiple-input-multiple-output (MIMO) radar sensors.

[0032] Short description of the drawings

[0033] Embodiments of the invention are illustrated in the drawings and explained in more detail in the following description.

[0034] Figure 1 shows a schematic representation of an antenna of a radar system in which the method according to the invention is used.

[0035] Figure 2 shows a flow diagram of an embodiment of the method according to the invention.

[0036] Embodiments of the invention Figure 1 shows an antenna of a radar system RS (not shown in detail), in which the method according to the invention is applied. The antenna has a plurality of antenna elements Rxl, Rx2, Rx3, Rx4, Txl, Tx2, Tx3 and Tx4, which emit a radar signal as transmitting antenna elements Txl, Tx2, Tx3 and Tx4 and receive a reflected radar signal as receiving antenna elements Rxl, Rx2, Rx3, Rx4. This is therefore a MIMO antenna (multiple input multiple output). Furthermore, transceivers T1 and T2 are provided, to which the antenna elements Rxl, Rx2, Rx3, Rx4, Txl, Tx2, Tx3 and Tx4 are connected. A first system-on-a-chip SoCl (also referred to as system-on-chip, SoC) consists of a first transceiver TI and two connected transmit antenna elements Txl, Tx2 and two connected receive antenna elements Rxl, Rx2.A second system-on-a-chip (SoC2) consists of a second transceiver (T2) and two connected transmit antenna elements (Tx3, Tx4) and two connected receiver antenna elements (Rx3, Rx4). The transmit antenna elements (Tx1, Tx2) form a first antenna group with the receive antenna elements (Rx1, Rx2) of the first system-on-a-chip (SoC1), and the transmit antenna elements (Tx3, Tx4) form a second antenna group with the receive antenna elements (Rx3, Rx4) of the second system-on-a-chip (SoC2). The two system-on-a-chips (SoC1, SoC2) are symmetrically configured and designed in a consistent manner. The geometry (G) of the antenna groups is known and stored in the specification of the radar system (RS) (see Figure 2). In addition, the transceiver hierarchy of the transceivers TI and T2 of the radar system RS is known.

[0037] Figure 2 shows a flowchart of an embodiment of the method according to the invention. Initially, the radar system RS carries out a measurement 10 of several targets in the surrounding area using the antenna elements Rxl, Rx2, Rx3, Rx4, Txl, Tx2, Tx3, Tx4. The measurement 10 is carried out with angular resolution, with the surrounding area being divided into angular cells by the antenna groups. The range and Doppler information 11 is then processed using fast Fourier transformation. Following this, the targets within the angular cells are detected 12. For each angular cell, the energy is determined, and a threshold for the noise energy is also determined. The determined energy is discriminated against the threshold for the estimated noise energy in each cell.In case several neighboring angle cells are above the threshold for the estimated noise energy, the local maxima of the measured complex data are determined for each combination of transmitters and receivers.

[0038] For self-calibration, the amplitude differences for the respective channels are compensated 20. For this purpose, the average received power across all signal amplitudes is calculated according to formula 2:

[0039] (Formula 2)

[0040] |x| represents the mean of the amplitudes and x mn the individual measured complex amplitudes. The summation is carried out over the transmitters m = 1...M and the receivers n = 1...N.

[0041] The deviation of the amplitude from the average received power is then calculated for each channel according to formula 3:

[0042] (Formula 3)

[0043] The deviation is then used to compensate for an amplitude error for each channel 20.

[0044] A two-dimensional linear regression is performed for each antenna group of the system-on-a-chip SoCl, SoC2 21. The regression plane is estimated for which the mean square difference between the phase measurement values ​​of the channels of the antenna groups is minimal. The two-dimensional regression uses the geometric data G of the antenna groups. For each channel, the difference between the measured phase value and the regression plane is calculated. This yields the intra-group phase correction value for the respective antenna group. If the phase error is angle-dependent, each angle is calibrated using its own target 22. Intermediate angles for which no separate target is available are interpolated 23. If the phase error is not angle-dependent, the measured values ​​of all targets for each channel are averaged 24.This is now repeated for all antenna groups so that intragroup phase correction values ​​Kintra are available for all (virtual) channels.

[0045] In addition, the distance between the two calculated regression planes of the two antenna groups of the first system-on-a-chip (SoCl) and the second system-on-a-chip (Son-a-Chip) is calculated 25, applying modulo 2n. This yields the intergroup phase correction value Kinter for the two antenna groups.

[0046] Finally, the control vectors of each channel are compensated 26 with the intragroup phase correction value Kintra and the intergroup phase correction value Kinter.

[0047] After self-calibration, further evaluations such as angle estimation 13 can be carried out from the reflex lists L.

Claims

Claims 1. Method for self-calibration of a radar system (RS) with at least two antenna groups, to which at least one transmission channel and one reception channel are assigned, characterized by the following steps: Measuring (10) several targets by the radar system (RS); For each antenna group, processing the range and Doppler information (11) and detection (12) of the targets to obtain reflex lists (L) with complex amplitudes for each target; Compensating (20) the amplitude differences for the respective channels; by means of two-dimensional linear regression (21), estimating a regression plane for which the mean square distance of the phase measurements of the channels of the antenna groups is minimal; Calculating the difference between the measured phase value and the regression plane for each channel to obtain an intragroup phase correction value (ntra); Calculating (25) the distance between two regression planes of different antenna groups, with modulo 2TT, to obtain an intergroup phase correction value ( nter); Compensating (26) the control vectors of each channel with the intragroup phase correction value (ntra) and the intergroup phase correction value (Knter).

2. Method according to claim 1, characterized in that when calculating the intragroup phase correction value (Kntra), in the case that there is no angular dependence of the phase error, the measured values ​​of all targets are averaged (24) for each channel.

3. Method according to claim 1, characterized in that when calculating the intragroup phase correction value (Kntra), in the case that the phase error is angle-dependent, each angle has its own Target is calibrated (22). Method according to claim 3, characterized in that intermediate angles for which no specific target is available are interpolated (23). Method according to one of the preceding claims, characterized in that the processing of the range and Doppler information (11) is carried out by a fast Fourier transformation. Method according to one of the preceding claims, characterized in that the detection (12) of the targets is carried out by the following steps: - Discrimination of the energy in each angle cell against a threshold for the estimated noise energy; - Determining the local maxima if several adjacent angle cells are above the threshold value. Method according to one of the preceding claims, characterized in that the compensation (20) of the amplitude differences is carried out by the following steps: Calculate the average received power; Calculating the deviation of the amplitude of each channel from the average received power; Compensate the amplitude based on the deviation.

Images