Method and apparatus for processing a Doppler distance matrix, and radar system

Abstract

The present invention relates to the processing of a Doppler range matrix of a radar system. In order to obtain a detection threshold of the Doppler range matrix simply, efficiently and quickly, only a subset of the cells of the Doppler range matrix is selected, and the detection threshold is obtained according to the selected subset of the cells of the Doppler range matrix.

Description

Technical Field

[0001] The present invention relates to a method and a device for processing a Doppler - Range - Matrix. The present invention also relates to a radar system. Background Art

[0002] For the processing of radar signals, the measurement process is usually divided into a plurality of measurement intervals that are successive in time. In addition, within a measurement interval, the measurement can also be divided into a plurality of cells. Each cell is defined here by the following parameter: the parameter characterizes the radar signal to be detected there. In a coherent frequency - modulated continuous - wave radar, for example, the cells of the Doppler - Range - Matrix can be defined in one dimension by the distance to a potential scattering center and in another dimension by the relative velocity of the potential scattering center.

[0003] To detect a target object, a so - called Constant False Alarm Rate (CFAR) algorithm can be used in radar signal processing to extract the target object, and by means of this method, the average interference power can be estimated locally and adaptively. Here, an adaptive threshold must be calculated in the detection step, and based on this adaptive threshold, the cells with signal components from the scattering centers of the object to be detected can be separated from the cells with pure noise components. If the amplitude of a cell is higher than this threshold, the scattering center from the object is detected, and the corresponding distance and velocity are marked in the list of detected objects.

[0004] Document DE 196 37 010 A1 discloses a method for radar signal processing and in particular discloses the determination of the average interference power of the radar cells of a matrix of a radar system. Summary of the Invention

[0005] The present invention discloses a method, a device, and a radar system for processing a Doppler - Range - Matrix. Other advantageous embodiments can be derived from the description.

[0006] Therefore, it is provided that:

[0007] A method for processing the Doppler - Range - Matrix of a radar system, the method having a step of selecting a subset of the cells of the Doppler - Range - Matrix and a step of obtaining a detection threshold. In particular, the detection threshold can be obtained in the case of using the values of the selected cells of the Doppler - Range - Matrix. In addition, the method further includes a step of detecting the scattering centers in the Doppler - Range - Matrix. Here, in particular, the scattering centers can be detected in the case of using the obtained detection threshold.

[0008] In addition, it is provided that:

[0009] A device for processing a Doppler range matrix of a radar system, the device having a selection device, a processing device and a detection device. The selection device is designed to select a subset of the cells of the Doppler range matrix. The processing device is designed to determine a detection threshold using the values of the selected cells of the Doppler range matrix. The detection device is designed to detect the scattering centers of the Doppler range matrix using the determined detection threshold.

[0010] Final setting:

[0011] A radar system having a radar sensor, a signal processing device and a device for processing a Doppler range matrix according to the invention. The radar sensor is designed here to transmit and receive radar signals. The signal processing device is designed to process the radar signals of the radar sensor and to generate a Doppler range matrix.

[0012] Advantages of the invention:

[0013] The invention is based on the following recognition: The Doppler range matrix generated in an operating radar system usually has a very large number of individual cells. In order to detect the scattering objects in these Doppler range matrices, it is usually necessary to determine the noise level, which is used to distinguish the values of the cells with scattering centers from the values of the cells with only noise components. Due to the large number of cells in the range-Doppler matrix, a complete analysis of the values of all cells requires a high computational effort here.

[0014] Therefore, the idea of the invention is to achieve an efficient, fast but reliable determination of the detection threshold taking this recognition into account in order to distinguish the cells with scattering centers in the range-Doppler matrix from the cells with pure noise components.

[0015] For this purpose, it is provided according to the invention that not all cells of the range-Doppler matrix are used to determine the detection threshold, but rather the detection threshold is determined only using a subset of the cells in the range-Doppler matrix. In this way, the computational effort for determining the detection threshold can be reduced. As a result, the required computing power can be reduced and, moreover, the time required for determining the detection threshold can be shortened. In this way, a rapid and efficient determination of the detection threshold for the range-Doppler matrix can be achieved with relatively inexpensive hardware.

[0016] As already mentioned at the beginning, the Doppler distance matrix involves a matrix with a large number of individual cells. Here, each cell of these two-dimensional matrices corresponds to a representative signal amplitude (of a distance and velocity pair within the radar sensor resolution). In other words, one dimension of these matrices represents the distance of the object to the radar sensor, while the other dimension represents the relative velocity between the object and the radar sensor. Here, the acquisition of these Doppler distance matrices by the radar sensor is in principle known and will therefore not be elaborated in detail.

[0017] Due to the numerous system characteristics of the radar system, each cell of the Doppler distance matrix also has a noise component. In order to detect an object (which means a scattering center in the monitoring area of the radar sensor), for each cell of the Doppler distance matrix it is checked whether the value in the cell is higher than the noise level to be determined. However, for this purpose, a suitable threshold needs to be distinguished separately for the acquired Doppler distance matrix in order to distinguish between pure noise and the signal of the scattering center.

[0018] For this purpose, it is provided that in order to determine the detection threshold for distinguishing between noise and the signal of the scattering center, not all cells of the Doppler distance matrix are considered in their entirety, but only a small subset of the cells of the Doppler distance matrix is analyzed, and then a suitable detection threshold is determined from these subsets.

[0019] Preferably, the subset of the cells of the Doppler distance matrix used for determining the detection threshold should include as much as possible: cells from the complete area (for which the detection threshold is to be determined) of the Doppler distance matrix.

[0020] As will be elaborated in more detail below, numerous different selection methods can be considered for selecting the cells that are used for determining the detection threshold. For example, the cells can be selected randomly or by means of a pseudo-random algorithm. In addition, cells or a group of cells that have been determined previously and that are considered for determining the detection threshold can also be selected. If necessary, the selection of the cells used for determining the detection threshold can also be adapted taking into account the scattering centers that have been detected previously.

[0021] In order to detect the scattering centers in the monitoring area of the radar sensor and thus detect objects, the individual cells of the Doppler distance matrix can be compared with the determined detection threshold. Here, the scattering centers can be detected in those cells whose value exceeds the determined detection threshold. Accordingly, the cells of the Doppler distance matrix thus detected can be recorded in the list of the detected objects. In particular, the distance and velocity corresponding to the respective cell, and if necessary the value of the respective cell, can be recorded in the list of the detected objects.

[0022] According to one embodiment, the detection of the scattering centers includes: comparing the value of a cell in the Doppler distance matrix with the determined detection threshold. As described above, if the value in a cell of the Doppler distance matrix exceeds the value of the determined detection threshold, a scattering center can be detected. Conversely, if the value in a cell of the Doppler distance matrix is below the detection threshold, it can be considered that this only relates to a noise component, and thus the corresponding cell does not represent a scattering center. It should be understood here that the detection of the scattering centers in the Doppler distance matrix is not limited to the detection of a single cell. Instead, it can be determined for each cell in the Doppler distance matrix individually whether it represents a scattering center.

[0023] According to one embodiment, in order to determine the detection threshold, the noise threshold of the Doppler distance matrix can first be determined. Here, in particular, the noise threshold can be determined taking into account the values in a selected subset of the cells of the Doppler distance matrix. In addition, in order to determine the detection threshold, a predetermined offset can also be added to the noise threshold. In this way, an additional safety margin in the form of a predetermined offset can be added to the determined noise threshold in order to further improve the reliability of detecting the scattering centers in the Doppler distance matrix.

[0024] According to one embodiment, the selection of the subset of cells of the Doppler distance matrix respectively includes: selecting a group of a plurality of adjacent cells at a predetermined position. For example, a plurality of positions can first be determined in the Doppler distance matrix, which means that a plurality of individual cells are first selected in the Doppler distance matrix. Then, for each of these positions in the Doppler distance matrix, a group of adjacent cells can be selected. For example, a group of n×m cells (for example 5×5 cells) can be selected for each position. In this way, for each of the selected positions, a region can be selected, whereby outliers of individual cells can be compensated for if necessary.

[0025] According to one embodiment, determining the detection threshold can include: determining a local noise threshold for each group of cells. In other words, first a common (for example average) noise threshold is determined for a group of a plurality of adjacent cells. Subsequently, the detection threshold can be determined using the determined local noise threshold. It should be understood here that the local noise threshold of a group of cells can be determined in any way (for example by taking an average, calculating the median or in any other way).

[0026] In particular, the determination of the detection threshold can be calculated based on a plurality of previously determined individual noise thresholds by means of any suitable statistical method.

[0027] According to one embodiment, the method for processing the Doppler distance matrix can be used individually for the rows and / or columns of the Doppler distance matrix. For example, an individual detection threshold can be determined for all cells in a row and / or all cells in a column. In addition, it is of course also possible to determine a single detection threshold for all cells of the Doppler distance matrix.

[0028] According to an alternative embodiment, the same detection threshold can be used for all cells of the Doppler distance matrix. Here, a single detection threshold for all cells of the Doppler distance matrix can be determined particularly simply and efficiently, such that the detection threshold can be determined very quickly, simply, and efficiently based on a subset of the cells of the Doppler distance matrix.

[0029] According to one embodiment, the subset of cells of the Doppler distance matrix for which the detection threshold is determined comprises at most 1% of the cells of the Doppler distance matrix. The selected subset can in particular comprise at most five per thousand, two per thousand, one per thousand, or, if necessary, even fewer cells of the Doppler distance matrix.

[0030] As long as it makes sense, the above configurations and extensions can be combined with one another arbitrarily. Other configurations, extensions, and embodiments of the invention also include combinations of the features of the invention described above or below in connection with the examples that are not explicitly mentioned. Those skilled in the art will in particular also add various aspects as improvements or supplements to the basic form of the invention. Description of the Drawings

[0031] The following explains other features and advantages of the invention with reference to the drawings. Shown here are:

[0032] Figure 1 A schematic diagram showing a block diagram of a radar system having a device for processing a Doppler distance matrix according to one embodiment;

[0033] Figure 2 A schematic diagram showing a block diagram of a device for processing a Doppler distance matrix;

[0034] Figure 3 A schematic diagram showing the determination of a detection threshold;

[0035] Figure 4 A flowchart showing a method for processing a Doppler distance matrix according to one embodiment. Detailed Description

[0036] Figure 1Schematic diagram showing a block diagram of a radar system having a device 1 for processing a Doppler distance matrix. The radar system may include, for example, a radar sensor 3 that transmits a radar signal S via a transmitting antenna 31. The transmitted radar signal S may be partially reflected or scattered by an object 100. A part of the reflected radar signal may be received as a received signal E by a receiving antenna 32 of the radar sensor 3. The pre-processed received signal may be processed by a signal processing device 2 of the radar system. The signal processing device 2 may in particular generate a Doppler distance matrix M. The function of the radar sensor 3 and the generation of the Doppler distance matrix M may be performed in any (e.g., conventional) manner here, and thus will not be elaborated in detail here.

[0037] Depending on the resolution, the Doppler distance matrix M may have hundreds of thousands, possibly millions, or possibly even more cells. Each cell of the Doppler distance matrix M corresponds here to a specific distance - relative velocity combination.

[0038] To analyze and process the Doppler distance matrix M, the Doppler distance matrix M is provided to the processing device 1. The processing device 1 first determines a detection threshold. Subsequently, the processing device 1 may compare the values in the respective cells of the Doppler distance matrix M with the determined detection threshold. If the value of a cell is higher than the detection threshold, a scattering center can be detected for that cell, which means that an object can be detected for the corresponding distance - relative velocity combination.

[0039] Since the number of cells in the Doppler distance matrix M is very large, calculating the detection threshold here incurs a very high overhead when using all the cells of the Doppler distance matrix. To simplify the determination of the detection threshold, the processing device 1 may here use a simplified and efficient determination of the detection threshold, which will be elaborated in more detail below.

[0040] Figure 2 Schematic diagram showing a block diagram of a device 1 for processing a Doppler distance matrix M. The Doppler distance matrix M is provided at the input of the device 1. The generation of such a Doppler distance matrix M may be performed, for example, by the previously described radar system. First, a subset of the cells of the Doppler distance matrix M is selected in a selection device 11. For this purpose, for example, a predetermined number of cells of the Doppler distance matrix M may be randomly selected. For example, at most 1% or even less - for example, five per thousand, two per thousand, one per thousand, or possibly even less - of the cells of the Doppler distance matrix M may be selected.

[0041] In addition to a purely random selection or a selection with the aid of a pseudo-random algorithm, a pre-determined scheme can be used to select a subset of the cells of the Doppler distance matrix M. Whether the subset of cells is randomly selected or selected according to a pre-given scheme, for the selection of the subset of cells of the Doppler distance matrix M, the cells of the subset should be distributed as evenly as possible over the entire area to be analyzed and processed of the Doppler distance matrix M.

[0042] In addition to a fixed pre-given scheme for selecting the cells of the Doppler distance matrix M, an individualized subset can be determined for each Doppler distance matrix M. Furthermore, the selection of the cells can also be adapted according to previous detection results. For example, when selecting cells, in particular the following cells are avoided: in previous detections, scattering centers have been detected in these cells. Alternatively, a pre-determined number of the following cells can also be selected when selecting the subset: in these cells, scattering centers have been detected in previous detections.

[0043] To select a subset of the cells of the Doppler distance matrix M, in addition to selecting individual cells, a group of cells, in particular a group of adjacent cells, can also be combined at different positions respectively. For example, multiple positions within the Doppler distance matrix M can be selected randomly or according to a pre-given scheme, and then a group of adjacent cells is combined at each selected position. For example, m×n cells (such as 5×5 cells) can be combined into a group respectively. Then, either the values of the cells of these groups can be considered individually, or the values of the groups can be combined first, and then the combined values can be further processed for each group.

[0044] After a subset of the cells of the Doppler distance matrix M has been combined, for example, the detection thresholds of the Doppler distance matrix M can be determined in the processing device 12. In particular, these detection thresholds can be determined in the case of using the subset previously selected in the selection device 11 of the cells of the Doppler distance matrix M. For this purpose, for example, the individual values of the cells of the subset of the cells of the Doppler distance matrix M can be analyzed in order to determine the noise threshold from these values. Here, the determination of the noise threshold can be carried out in any suitable manner. For example, a statistical analysis can be implemented in order to identify the following threshold: this threshold marks the boundary between the noise values and the significant values (such as those representing scattering centers). For this purpose, for example, it can be considered that in most of the cells of the Doppler distance matrix (and thus in most of the selected cells of the subset) there are only noise values. If scattering centers are present in one or more of the cells of the subset of the cells of the Doppler distance matrix M, it can be considered that these cells only account for a relatively small share of the selected cells.

[0045] In addition, in order to further additionally improve the reliability of detecting the scattering center, a safety margin in the form of a predetermined offset or the like can also be added to the detected noise threshold described above when necessary. In this way, the detection threshold is obtained from the sum of the detected noise threshold and the safety margin in the form of the predetermined offset.

[0046] Then, the Doppler distance matrix M can be analyzed and processed using the obtained detection threshold to obtain one or more scattering centers in the Doppler distance matrix. For this purpose, the detection device 13 can compare the values in the respective cells of the Doppler distance matrix M with the obtained detection threshold. If the value in the cell of the Doppler distance matrix M is higher than the obtained detection threshold, then this cell can be considered to represent a scattering center. Correspondingly, these cells can be recorded in a hit (i.e., the detected object) list. For this purpose, for example, the distances and relative velocities represented by these cells can be recorded in the hit list. When necessary, the specific values of these cells can also be recorded together in the hit list.

[0047] Here, for analyzing and processing the Doppler distance matrix M, as described above, a common, previously obtained detection threshold is used for all cells of the Doppler distance matrix M. However, in principle, it is also possible to obtain individualized detection thresholds for each row or column separately. Correspondingly, the subset of cells of the Doppler distance matrix M used to obtain the detection threshold only relates to the cells of these rows or columns accordingly. In this way, individualized detection thresholds can be used for each distance or relative velocity respectively.

[0048] Figure 3 A schematic diagram showing a method for obtaining the detection threshold of the Doppler matrix M according to an embodiment is shown. In Figure 3 the upper region, the Doppler distance matrix M is schematically shown for this purpose. First, a subset of cells can be selected from this Doppler distance matrix M. As already described above, each cell can be selected individually here. Alternatively, a plurality of positions 50 in the Doppler distance matrix M can also be selected first. Then, for each of these positions 50, a group of adjacent cells 51 is selected. In this case, a common noise threshold can be obtained for each group 51 of cells first. For this purpose, for example, an average value, a median value, or any other suitable value can be constructed. In a further process, the values of the noise thresholds of these groups 51 can then be analyzed and processed separately to obtain the noise threshold. On the contrary, if the individual cells of the subset are considered individually, the values of these cells can also be considered directly for obtaining the noise threshold.

[0049] As Figure 3As shown in the lower region of, for example, a histogram or the like can be constructed in which the occurrence frequency N of each value P of the values in the cells of the Doppler distance matrix M or the occurrence frequency N of the noise threshold determined before the group 51 is combined.

[0050] Here, it can be considered that the values with high occurrence frequencies correspond to the cells without scattering centers. In other words, it can be considered that there are no scattering centers in the vast majority of the cells of the Doppler distance matrix (and thus in the vast majority of the cells of the selected subset of the Doppler distance matrix M), and thus these cells only have noise values.

[0051] Accordingly, the noise threshold R for the Doppler distance matrix M can be determined as a value slightly higher than the accumulated level value. This is shown by the dotted line in Figure 3 the lower region. As already described before, a safety margin in the form of an offset O can also be added to these noise thresholds for further bounding. Accordingly, the detection threshold D is obtained by the line shown as the dashed line in Figure 3 .

[0052] Figure 4 A schematic diagram showing a flowchart of a method for processing a Doppler distance matrix M according to an embodiment is shown. As already described above in connection with the device 1 for processing the Doppler distance matrix M, this method can in principle implement any step. Accordingly, the previously described device 1 can also have any components in order to implement the following steps described above.

[0053] In step S1, a subset of the cells of the Doppler distance matrix M is first selected. Subsequently, the detection threshold D can be determined in step S2. In particular, the detection threshold D is determined using the values of the selected cells of the Doppler distance matrix M. Then, the scattering centers in the Doppler distance matrix M can be detected in step S3, where the scattering centers are detected using the determined detection threshold D. In particular, the scattering centers can be detected in the cells of the Doppler distance matrix M that have values higher than the determined detection threshold.

[0054] In summary, the present invention relates to the processing of a Doppler distance matrix of a radar system. In order to simply, efficiently, and quickly determine the detection threshold of the Doppler distance matrix, only a subset of the cells of the Doppler distance matrix is selected, and the detection threshold is determined based on the selected subset of the cells of the Doppler distance matrix.

Claims

1. A method for processing a Doppler distance matrix (M), the method having the following steps: Selecting (S1) a subset of the cells of the Doppler distance matrix (M); Determining (S2) a detection threshold (D) using the values of the selected cells of the Doppler distance matrix (M); Detecting (S3) scatter centers in the Doppler distance matrix (M) using the determined detection threshold (D), wherein, a subset of the cells of the Doppler distance matrix (M) is randomly selected (S1) by means of a pseudo-random algorithm, wherein the cells of the subset are evenly distributed over the entire area to be analyzed and processed of the Doppler distance matrix (M), wherein detecting (S3) the scatter centers includes: comparing the value of each cell of the Doppler distance matrix (M) with the determined detection threshold (D), wherein the scatter centers are detected in the following cells: the value of the cell exceeds the determined detection threshold, wherein a plurality of positions (50) within the Doppler distance matrix (M) are randomly selected by means of a pseudo-random algorithm, and for each of the plurality of positions a group of a plurality of adjacent cells (51) is selected, wherein determining (S2) the detection threshold (D) includes: determining a local noise threshold for each group of cells (51), and determining the detection threshold (D) using the determined local noise threshold.

2. The method according to claim 1, wherein, determining (S2) the detection threshold (D) determines a noise threshold (R) using the values of the selected cells of the Doppler distance matrix (M), and calculates the detection threshold (D) from the sum of the noise threshold (R) and an offset (O).

3. The method according to any one of claims 1 to 2, wherein, the method is individualized for the rows and / or columns of the Doppler distance matrix (M).

4. The method according to any one of claims 1 to 2, wherein, the same detection threshold (D) is used for all rows of the Doppler distance matrix (M).

5. The method according to any one of claims 1 to 4, wherein, the selected subset includes at most one percent of the cells of the Doppler distance matrix (M).

6. A device (1) for processing a Doppler distance matrix (M), the device having: selecting means (11) designed to select a subset of the cells of the Doppler distance matrix (M); processing means (12) designed to determine a detection threshold (D) using the values of the selected cells of the Doppler distance matrix (M); detecting means (13) designed to detect scatter centers in the Doppler distance matrix (M) using the determined detection threshold (D), wherein, a subset of the cells of the Doppler distance matrix (M) is randomly selected by means of a pseudo-random algorithm, wherein the cells of the subset are evenly distributed over the entire area to be analyzed and processed of the Doppler distance matrix (M), Among them, detecting the scattering center includes: comparing the value of each cell of the Doppler distance matrix (M) with the detected threshold (D) to be obtained, wherein the scattering center is detected in the following cells: the value of the cell exceeds the detected threshold to be obtained. Among them, a plurality of positions (50) within the Doppler distance matrix (M) are randomly selected by means of a pseudo-random algorithm, and for each of the plurality of positions, a set of a plurality of adjacent cells (51) is selected, wherein obtaining the detected threshold (D) includes: obtaining a local noise threshold for each set of cells (51), and obtaining the detected threshold (D) in the case of using the obtained local noise threshold.

7. A radar system, the radar system having: A radar sensor (3), the radar sensor being designed to transmit and receive radar signals; A signal processing device (2), the signal processing device being designed to process the radar signals of the radar sensor (3) and to generate a Doppler distance matrix (M); The device (1) for processing the Doppler distance matrix (M) according to claim 6.

Images