Methods for self-calibration of radar systems

Abstract

The present invention relates to a method for self-calibration of a radar system (RS) comprising at least two antenna groups, each assigned with at least one transmit channel and at least one receive channel. First, the radar system (RS) measures (10) a number of targets. Then, for each antenna group, range and Doppler information processing (11) and target detection (12) are performed to obtain, for each target, a reflection list (L) with complex amplitudes. Then, for each channel, compensation (20) for amplitude differences is performed. Then, by two-dimensional linear regression (21), a regression plane is estimated that minimizes the mean-square distance of the phase measurements of the channels of the antenna group. For each channel, the difference between the measured phase value and the regression plane is calculated to obtain an intra-group phase correction value (K intra ) is obtained. Furthermore, the distance between two regression planes of different antenna groups is calculated modulo 2π (25) to obtain the inter-group phase correction value (K inter ) is obtained. Finally, the intra-group phase correction value (K intra ) and the intergroup phase correction value (K inter ) are used to compensate (26) the control vectors for each channel.

Description

Technical Field

[0001] The present invention relates to a method for self-calibration of a radar system comprising at least two antenna groups. At least one transmit channel and at least one receive channel are assigned to each antenna group.

Background Art

[0002] Radar systems are used to measure the distance, relative velocity, and azimuth and elevation angles of objects. For angle estimation, an antenna group, also called an antenna array, which can be used for both transmission and reception, is typically used. By digital beamforming, the angle of incidence of the reflected plane wave generated by a target at a distance is captured. Regarding beamforming, the phase difference of the reflected wave is evaluated across a plurality of receive channels. The received signals of the antenna group can be processed by conversion as if they were measured by a virtual receiver. Conventionally, the evaluation has been performed by previously stored control vectors, which are applied in various ways. For example, in a Bartlett beamformer, the expected phase differences for various angles of incidence are stored in the control vectors and correlated. Alternatively, a model-based estimation based on the control vectors can also be performed.

[0003] Conventionally, the control vectors are measured over an angle (one or more angular regions) for each individual sensor during a single end-of-line calibration and then stored in non-volatile memory. Thereby, phase errors (phase offsets) that can occur due to various influences, such as manufacturing tolerances of antennas or supply lines, or the interaction of electromagnetic waves with the radome, housing, or circuit board, are taken into account.

[0004] The antenna group further divides the surrounding environment into angular cells, allowing them to be intercepted individually. This enables the separation of targets with the same range and radial velocity based on their angular position. Here, each radar beam of the radar system acts as a spatial filter.

[0005] Special functions in vehicle automation require angle measurement with superior accuracy and resolution. To achieve this, antenna groups with large apertures are used. To control the corresponding transceivers, the transceivers are typically grouped and cascaded, thereby creating a transceiver hierarchy. To perform coherent measurement of the incident angle of the reflected wave, all transceivers are preferably connected to a reference oscillator. For this purpose, a high-frequency line is typically used to distribute the signal from the reference oscillator to the transceivers. However, the larger the aperture, the longer the high-frequency line becomes.

[0006] Currently, patch antennas are commonly used as radar antennas in the automotive sector. Patch antennas are mounted on the surface of a circuit board and powered by a stripline. Therefore, the size of the circuit board increases in proportion to the size of the aperture. Consequently, larger circuit boards are more likely to have uneven temperature distribution on the circuit board, generated by internal and / or external influencing factors. Examples of internal influencing factors include strong heating from various component groups. Examples of external influencing factors include partial shielding of the circuit board and heating from adjacent component groups and / or airflow.

[0007] Factors such as uneven temperature distribution and component aging affect radar system measurements by altering previously calibrated phase errors. If angle estimation continues using the initially stored control vectors, the performance of angle estimation degrades. Specifically, side lobes increase in the angular spectrum, the main lobe is weakened and broadened, and its position changes (beam pointing error). As a result, the dynamic range, accuracy, and resolution of angle estimation decrease.

[0008] Self-calibration of radar systems is known from 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. Here, it is assumed that the wavefront of the reflected wave at the antenna is flat. This assumption is sufficiently satisfied only when the target is in the far field. The boundary with respect to the far field is typically defined by Equation 1, relating the aperture length D and wavelength λ.

[0009]

number

[0010] Assuming the target is in a far field, a one-dimensional linear regression is performed from the received phase at the antenna elements of the receiving antenna group, thereby estimating the most likely wavefront profile. From this, the phase error to be corrected is calculated. The phase error estimated from the target is applied for each target angle without constraints of generality. Here, it is not important whether the radar system has a radome or is located behind, for example, a bumper.

[0011] A further prerequisite for self-calibration according to the above literature is that the antenna elements are arranged at equal intervals along one dimension. Here, the following arrangement is given: the antenna elements are arranged horizontally to calibrate the azimuth control vector, and the antenna elements are arranged vertically to calibrate the elevation control vector.

[0012] Generally, self-calibration can be used for various modulation methods. Today, typical transmission frequencies are 24 GHz or 77 GHz, and the maximum demonstrable bandwidth is less than 4 GHz, particularly around 0.5 GHz.

[0013] Modern radar systems in the automotive sector typically use FMCW modulation (Frequency Modulated Continuous Wave Radar) with fast ramps (fast chirp modulation), where multiple linear frequency ramps of the same slope are transmitted sequentially. The mixing of the current transmitted and received signals generates a low-frequency signal whose frequency (called the beat frequency) is proportional to the distance. The system is usually designed so that the beat frequency component caused by the Doppler frequency is negligibly small. The distance information obtained from the beat frequency is nearly unique, and the Doppler shift can then be determined by observing the temporal change in the phase of the complex distance signal across the entire ramp. Distance and velocity are determined independently of each other, usually using a two-dimensional Fourier transform. The angle estimation described above is performed downstream of the distance and velocity estimation. [Prior art documents] [Non-patent literature]

[0014] [Non-Patent Document 1] 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 [Overview of the Initiative] [Means for solving the problem]

[0015] A method for self-calibration of a radar system (RS) is proposed, comprising at least two antenna groups, each assigned at least one transmit channel and at least one receive channel. The geometry of the antenna groups, i.e., the positions of the transmit and receive antennas, is known. Furthermore, the transceiver hierarchy of the transceivers used to form the antenna groups, whose elements undergo similar phase errors, is known. These data are part of the specification of the radar system.

[0016] First, multiple targets are measured by the radar system. For this purpose, a transmitting antenna group emits electromagnetic waves in the direction of the targets. Then, the electromagnetic waves reflected from the targets are received by a receiving antenna group and evaluated using digital beamforming. The measurement is performed using an angle-resolved method, where the surrounding environment is divided into angular cells by the antenna group.

[0017] For each antenna group, range and Doppler information is processed (also called range-Doppler processing). This processing is preferably performed by a two-dimensional fast Fourier transform, but can also be performed by other algorithms known per se. Thereafter, detection of targets within an angular cell is performed. For this purpose, preferably the following steps are performed: For each angular cell, energy is calculated. Further, a threshold regarding the estimated noise energy of such an angular cell is estimated. The calculated energy is compared with the threshold regarding the noise energy, and thus, the calculated energy in each cell is distinguished from the estimated noise energy. If several adjacent angular cells exceed the threshold regarding the estimated noise energy, a local maximum of the measured complex data is determined for each combination of transmitter and receiver. The reflection list is obtained by the above steps, and the reflection list stores, for each target, the range, relative velocity, and complex amplitude measured in each respective virtual channel.

[0018] First, for each channel, the amplitude difference is compensated. For this purpose, preferably the following steps are performed. The average received power over all signal amplitudes is calculated, where in particular Equation 2 is used.

[0019]

Number

[0020] Here,

[0021]

Number

[0022] represents the average value of the amplitude,

[0023]

Number

[0024] This represents the individual complex amplitudes measured. These are summed over the transmitter m=1···M and the receiver n=1···N. Subsequently, the amplitude deviation relative to the average received power is calculated for each channel. In particular, the deviation is calculated by the quotient given by Equation 3.

[0025]

number

[0026] Next, this deviation is used to compensate for the amplitude error for each channel. This step enables calibration of the received power. Next, intra-group calibration is performed for each antenna group. For this purpose, a two-dimensional linear regression is performed for each antenna group. Here, the regression plane that minimizes the mean squared distance of the phase measurements of the antenna group's channels is estimated. In the two-dimensional regression, the geometric data of the antenna groups, which is known as part of the radar system specifications as mentioned at the beginning, is used. Here, for each channel, the difference between the measured phase value and the regression plane is calculated. This gives the intra-group phase correction value for each antenna group. This step is then repeated for all antenna groups, thereby obtaining the intra-group phase correction value for all (virtual) channels.

[0027] If the phase error is not angle-dependent, for example, if the phase field is based on the aging of the radar system, all target measurements can be averaged for each channel. This yields an average gain. If the phase error is angle-dependent, for example, if the radar system is located behind a bumper, each angle can be calibrated with respect to a specific target. Preferably, intermediate angles without a specific target may be interpolated. Here, the angles in which targets exist must be close enough to each other that the sampling theorem is preserved with respect to the phase error. By performing multiple measurement cycles sequentially, targets in angle cells that were previously only interpolated or for which phase error estimation was not possible can be captured.

[0028] Next, inter-group calibration is performed for at least two different antenna groups to determine the phase error between the different antenna groups. For each antenna group, the regression plane was calculated as described above. Here, the distance between the two regression planes of the two different antenna groups is calculated, and modulo 2π is applied. This gives the inter-group phase correction value for the two antenna groups. Once the regression planes are compensated with the inter-group phase correction value, all the regression planes of the antenna groups overlap each other.

[0029] Finally, the control vectors for each channel are compensated using intra-group and inter-group phase compensation values. Angle estimation can then be performed according to known methods.

[0030] Due to its two-dimensional regression, this method can generally be applied to all geometric shapes of antenna groups, without restrictions in terms of orientation or distance. Antenna groups do not necessarily need to be positioned along horizontal or vertical lines and can be freely positioned. Furthermore, antenna elements do not need to be positioned equidistant from each other and can be located at any distance from one another. In addition, the phase errors for each channel do not need to be evenly distributed, and different average phase errors are possible for each antenna group. This method can also be used for multiple input multiple output radar sensors (MIMO).

[0031] Exemplary embodiments of the present invention are shown in the drawings and described in more detail below. [Brief explanation of the drawing]

[0032] [Figure 1] This is a schematic diagram of an antenna in a radar system in which the method according to the present invention is used. [Figure 2] This is a flowchart illustrating an embodiment of the method according to the present invention. [Modes for carrying out the invention]

[0033] Figure 1 shows an antenna for a radar system RS (not shown in detail) to which the method according to the present invention is applied. The antenna comprises multiple antenna elements Rx1, Rx2, Rx3, Rx4, Tx1, Tx2, Tx3, and Tx4, which transmit radar signals as transmitting antenna elements Tx1, Tx2, Tx3, and Tx4 and receive reflected radar signals as receiving antenna elements Rx1, Rx2, Rx3, and Rx4. In other words, this is a MIMO (Multiple Input Multiple Output) antenna. Furthermore, transceivers T1 and T2 are provided, to which the antenna elements Rx1, Rx2, Rx3, Rx4, Tx1, Tx2, Tx3, and Tx4 are connected. The first system-on-chip SoC1 (also called a system-on-chip or SoC) consists of a first transceiver T1, two transmitting antenna elements Tx1 and Tx2 connected to the first transceiver T1, and two receiving antenna elements Rx1 and Rx2 connected to the first transceiver T1. The second system-on-chip SoC2 consists of a second transceiver T2, two transmitting antenna elements Tx3 and Tx4 connected to the second transceiver T2, and two receiving antenna elements Rx3 and Rx4 connected to the second transceiver T2. The transmitting antenna elements Tx1 and Tx2, together with the receiving antenna elements Rx1 and Rx2 of the first system-on-chip SoC1, form a first antenna group, and the transmitting antenna elements Tx3 and Tx4, together with the receiving antenna elements Rx3 and Rx4 of the second system-on-chip SoC2, form a second antenna group. The two system-on-a-chip SoCs, SoC1 and SoC2, are configured symmetrically and designed similarly. The antenna group geometry G is known and stored in the specifications of the radar system RS (see Figure 2). Furthermore, the transceiver hierarchy of transceivers T1 and T2 of the radar system RS is known.

[0034] Figure 2 shows a flowchart of an embodiment of the method according to the present invention. First, the radar system RS uses antenna elements Rx1, Rx2, Rx3, Rx4, Tx1, Tx2, Tx3, and Tx4 to perform measurements 10 of multiple targets in the surrounding environment. The measurements 10 are performed using the angle-resolved method, and the surrounding environment is divided into angle cells by the antenna group. Subsequently, range and Doppler information processing 11 is performed using the Fast Fourier Transform. Next, target detection 12 is performed within the angle cells. For each angle cell, the energy is calculated, and a threshold for noise energy is also calculated. The calculated energy is distinguished for each cell against the threshold for estimated noise energy. If several adjacent angle cells exceed the threshold for estimated noise energy, the local maximum value of the measured complex data is determined for each combination of transmitter and receiver.

[0035] For self-calibration, the amplitude difference between each channel is compensated. For this purpose, the average received power across all signal amplitudes is calculated according to Equation 2.

[0036]

number

[0037] Here,

[0038]

number

[0039] This represents the average value of the amplitude.

[0040]

number

[0041] This represents the individual complex amplitudes measured. These are summed over the transmitter m=1···M and the receiver n=1···N. Subsequently, for each channel, the amplitude deviation relative to the average received power is calculated according to Equation 3.

[0042]

number

[0043] Next, the amplitude error is compensated for each channel using the deviation. For each antenna group of System-on-Chip SoC1 and SoC2, a two-dimensional linear regression 21 is performed. Here, in each case, the regression plane that minimizes the mean squared distance of the phase measurements of the channels in the antenna group is estimated. Geometric data G of the antenna group is used in the two-dimensional regression. Here, for each channel, the difference between the measured phase value and the regression plane is calculated. This gives an intra-group phase correction value for each antenna group. If the phase error is angle-dependent, each angle is calibrated 22 with respect to a unique target. Intermediate angles without a unique target are interpolated 23. If the phase error is not angle-dependent, for each channel, it is averaged 24 over all target measurements. Here, this is repeated for all antenna groups, thereby giving an intra-group phase correction value K for all (virtual) channels. intra This can be obtained.

[0044] Furthermore, the distance between the two calculated regression planes of the two antenna groups of the first system-on-chip SoC1 and the second system-on-chip is calculated, and modulo 2π is applied. This gives the inter-group phase correction value K for the two antenna groups. inter This can be obtained.

[0045] Finally, the group phase correction value K intra and inter-group phase correction value K inter Compensation 26 of the control vectors for each channel is performed using the above. After self-calibration, further evaluations such as angle estimation 13 can be performed from the reflection list L.

Claims

1. A method for self-calibrating a radar system (RS) having at least two antenna groups to which at least one transmit channel and at least one receive channel are assigned, The steps include measuring (10) multiple targets using the radar system (RS), For each antenna group, the steps include processing the range and Doppler information of the target (11) and detecting the target (12) to obtain a reflection list (L) having a complex amplitude for each target, The steps include compensating for the amplitude difference (20) for each channel, The steps include: using two-dimensional linear regression (21), estimating the regression plane that minimizes the mean squared distance of the phase measurements of the channels in the antenna group; For each channel, the difference between the measured phase value and the regression plane is calculated to obtain the group phase correction value (K ｉｎｔｒａ The steps to obtain ) and The distance between two regression planes of different antenna groups is calculated modulo 2π (25), and the inter-group phase correction value (K ｉｎｔｅｒ The steps to obtain ) and The group-internal phase correction value (K ｉｎｔｒａ ) and inter-group phase correction value (K ｉｎｔｅｒ The steps include compensating the control vector of each channel (26) using ) and A method characterized by the following.

2. The group-internal phase correction value (K ｉｎｔｒａ The method according to claim 1, characterized in that when calculating the phase error, if the phase error is angle-dependent, each angle is calibrated with respect to a specific target (22).

3. The method according to 2, characterized in that intermediate angles without a specific target are interpolated (23).

4. The method according to claim 1, characterized in that the processing (11) of the range and Doppler information is performed by the Fast Fourier Transform.

5. The detection (12) of the target is The steps include distinguishing the energy at each angle cell against a threshold for the estimated noise energy, If multiple adjacent angle cells exceed the threshold, the step of determining the maximum value is as follows: The method according to claim 1, characterized in that it is carried out by [a specific method].

6. The compensation (20) for the amplitude difference is, The steps include calculating the average received power and A step of calculating the deviation of the amplitude of each channel from the average received power, A step of compensating the amplitude according to the deviation, The method according to claim 1, characterized in that it is carried out by [a specific method].

Images