Method for controlling a coherent cooperative radar sensor network

Abstract

The present invention relates to a method for controlling a coherent cooperative radar sensor network comprising multiple radar sensors (11, 12, 13), at least two of which operate coherently. Sensor data (SD1, SD2, SD3) are divided into data to be coherently evaluated (KD) and data to be non-coherently evaluated (NKD) according to the evaluation type. The data to be coherently evaluated (KD) and the data to be non-coherently evaluated (NKD) are transmitted to different evaluation units (110, 120, 130; 31, 32, 33), which then perform the respective evaluations. The individual evaluations are combined to generate an overall evaluation.

Description

[Technical Field]

[0001] The present invention relates to a method for controlling a coherent cooperative radar sensor network, and further to a coherent cooperative radar sensor network implementing this method. [Background technology]

[0002] DE 10 2015 224 787 A1 describes a coherent cooperative radar sensor network consisting of at least two radar sensors. These radar sensors are synchronized with each other by exchanging data or via signals. Each radar sensor transmits information about its radar signal to a processing device. The processing device may be external or integrated into one of the radar sensors. The processing device processes the received information and preferably determines both bistatic and monostatic distances to an object.

[0003] DE 10 2019 220 238 A1 discloses a coherent cooperative radar sensor network consisting of at least two radar sensors and a method for calibrating the same, in which a phase control signal is transmitted between the radar sensors and a radar signal is transmitted based on the phase control signal, and an evaluation unit evaluates the received signals of the radar sensors.

[0004] The evaluation of radar signals is traditionally performed centrally in an evaluation unit. This applies to both coherent and non-coherent data. The evaluation unit may be part of one radar sensor, and the data of the remaining radar sensors are then transmitted to the evaluation unit.

[0005] It is further known for non-coherent radar sensor networks to evaluate non-coherent data in a distributed manner with these radar sensors. Summary of the Invention

[0006] A method is proposed for controlling a coherent cooperative radar sensor network. The coherent cooperative radar sensor network comprises a plurality of interconnected and cooperative radar sensors. At least two sensors, preferably all sensors, operate coherently. However, non-coherent cooperative sensors may also be contemplated.

[0007] It is envisaged that the sensor data to be evaluated are divided and evaluated in different evaluation units. These evaluation units may be physical computing devices or processors or may be realized as virtual cores or software blocks. Depending on the evaluation type, the sensor data are divided into data to be evaluated coherently (hereinafter also referred to as coherent data) and data to be evaluated non-coherently (hereinafter also referred to as non-coherent data). For this purpose, a (two-dimensional) spectrum is preferably calculated. The division is based on the spectrum. Here, ranges to be evaluated coherently or non-coherently are advantageously calculated. Complex spectral samples are then distributed according to these ranges. In frequency-modulated continuous wave radar (FMCW), the division can also be performed by filtering based on a time signal. Coherent data is obtained from a high-pass filter, and non-coherent data is obtained from a low-pass filter. The data to be evaluated coherently and non-coherently are transmitted to different evaluation units. The data then reside in the evaluation units, which then perform the respective evaluations of the data. Thus, the non-coherent data is evaluated by at least one evaluation unit and separately the coherent data is evaluated by at least one further evaluation unit, and the individual evaluations are then combined to produce an overall evaluation, which is performed in the evaluation unit.

[0008] The partitioning distributes the computational effort among multiple evaluation units, allowing the individual evaluations to be performed in parallel, thus reducing the amount of computation required by each evaluation unit or improving overall performance.

[0009] When evaluating non-coherent data, for example, vector velocity is acquired and / or a joint target list is created (position aggregation). When evaluating coherent data, phase and / or frequency synchronization can be performed via the antenna array and / or angle estimation can be performed. See DE102019220238A1.

[0010] Preferably, the allocation of evaluation units to data is based on the expected computational effort or the expected required computational volume for data of each evaluation type, in order to equalize the computational effort required for evaluating each data set between the evaluation units. For example, evaluation of coherent data requires more computational volume than evaluation of non-coherent data. Therefore, inter alia, the number of evaluation units and / or the computational volume of the evaluation units are preferably selected depending on the evaluation type.

[0011] Preferably, to achieve a constant data rate, it is intended that the divided data have the same amount of data. This is particularly important for multiple cooperative sensors (e.g., four or more sensors). Here, it would be possible to subdivide the coherent and / or non-coherent components again. Preferably, the threshold is dynamically adapted to the data.

[0012] In the overall evaluation, the data from the individual evaluations for each target are combined in the evaluation unit, so that in the end, all antenna combinations of the transmitting and receiving antennas of all sensors are evaluated together, which is necessary, for example, for coordinated angle calculations.

[0013] Further division of the sensor data to be evaluated may be contemplated: The sensor data may be divided according to one or more of the following criteria related to the radar signal used:

[0014] Distance Range - Here you can determine the distance threshold for splitting. Doppler / Velocity Range - Velocity can be calculated from the Doppler shift, so it can be used directly as a reference or a velocity threshold can be set.

[0015] Angle range. These criteria represent meaningful cut levels for the sensor data. A combination of these criteria can also be used for the segmentation. In particular, the segmentation into coherent and non-coherent data and the segmentation into distance ranges can be advantageously combined, since the coherent angle evaluation only becomes meaningful after a certain separation distance, where the far-field assumption applies.

[0016] The threshold value can be optimized depending on the data rate, the amount of calculation, the calculation operation, etc. Preferably, a common separation level is determined before the information exchange. For this purpose, for example, a handshake protocol can be performed between the sensors. For example, a minimum threshold value can always be used for the division. Preferably, in order to achieve a constant data rate, it is intended that the cuts have the same amount of data. This is particularly important when the division results in certain ranges containing more targets than other ranges. Preferably, the threshold value is dynamically adapted to the data.

[0017] The data for different criteria are then transmitted to different evaluation units, which then perform evaluations of the data respectively. The individual evaluations are then combined to generate an overall evaluation, which is then performed in the evaluation units. The division distributes the computational effort among multiple evaluation units, allowing the individual evaluations to be performed in parallel, thus reducing the amount of computation required in each evaluation unit.

[0018] To split the sensor data, the sensor data can be transmitted to a data distributor unit, which splits the sensor data as described above and then distributes the sensor data to the evaluation units accordingly. The data distributor unit can be part of one or more sensors or can be configured as a stand-alone unit.

[0019] Today, radar sensors often already have an evaluation unit. In particular, data can be distributed to the evaluation unit located in the radar sensor. For this purpose, the aforementioned data distributor unit can preferably be used. A bidirectional connection to the sensor is then required. The evaluation unit located in the radar sensor then performs the evaluation of the data. This has the advantage that the evaluation can be performed directly in the radar sensor, and no central computing device is required.

[0020] Furthermore, compared to central evaluation, the overall data rate between the evaluation units is reduced: each part of the data remains in the sensor or its evaluation unit until it is fully evaluated. If the data load between the N evaluation units is the same, the following relationship applies to the data rates in both cases:

[0021]

number

[0022] As an example for three sensors, the overall data rate is reduced by a factor of three compared to central evaluation. Preferably, a data buffer is provided, since the data in the evaluation unit of the radar sensor can only be discarded or overwritten after a complete exchange. Sensordaten If is homogeneous, buffer D Puffer The size of can be estimated by the following relationship:

[0023]

number

[0024] In the example for three evaluation units, the size of the buffer is thereby reduced by more than a factor of 3 compared to a central evaluation unit. In the data distributor unit described above, in addition to splitting the data, buffering can also be performed.

[0025] Alternatively or additionally, a central computing device with multiple evaluation units may be provided. The data is transmitted to the central computing device and split there among the various evaluation units. The splitting can be performed by the aforementioned data splitter unit, which may further be designed as a separate unit within one or more sensors or as part of the central computing device. Multiple evaluation units within the central computing device then perform the evaluation of the data. Here, the data of each sensor can be evaluated independently of each other until the angle estimation. In particular, the multiple evaluation units are realized as different processors of the computing device, as virtual cores, or as software blocks. Preferably, the evaluation units of the central computing device access a central memory, so that the data is available to each evaluation unit. Alternatively, a memory may be provided for each evaluation unit or for groups of evaluation units.

[0026] If there is both an evaluation unit at the sensor and a central computing device with multiple evaluation units, it is preferably intended that the sensor data to be evaluated non-coherently be transmitted to the evaluation unit of the non-coherent laser sensor, since this requires less computational effort. The sensor data to be evaluated coherently can then be transmitted to the central computing device, where the more powerful evaluation unit takes over the evaluation with greater computational effort.

[0027] According to one aspect, the sensor data can be transmitted as raw data from the sensor. This allows the use of simple sensor heads that deliver raw data, such as analog-to-digital converter data. Optionally, the raw data can be decimated before transmission. Alternatively, Fourier transformed data, such as range FFT data, Doppler FFT data, and / or range-Doppler FFT data, can be transmitted. This allows the use of simple, inexpensive radar sensors that do not require a pre-processing unit for the raw data and a simple pre-processing unit that performs a Fourier transform for the Fourier transformed data. When a central computing device is used for evaluation, the radar sensor does not need to include an evaluation unit. Furthermore, the raw data or Fourier transformed data conveys all information about all targets. This allows for a high data rate during transmission, but no information is lost. When only raw data is transmitted, measurements typically performed during pre-processing, such as separation and / or Doppler evaluation using a fast Fourier transform, can be performed during evaluation.

[0028] During joint evaluation, noncoherent integration (NCI) can be performed based on multiple sensors. NCI calculations can be performed on both coherent and noncoherent data. Preferably, joint NCI calculations are performed for targets with far-field conditions, since for these targets, it can be assumed that the target can be detected by multiple sensors within the same range-velocity cell. For close ranges (noncoherent ranges), NCI calculations are preferably performed individually based on data from individual sensors or bistatic measurements. Individual evaluation carries the risk of the target not being recognized by the individual sensors. Targets missing from individual sensors or bistatic measurements can be detected later by a central NCI calculation before the data is finally discarded. This achieves a higher integration gain for downstream target detection compared to individual evaluation, e.g., with a constant false alarm probability (CFAR). As a result, the sensitivity and range of the cooperative sensor network are improved. Finally, further signal processing steps typical for each data type, such as angle estimation and / or velocity estimation, as well as non-coherent processing steps (e.g., vector velocity calculation and / or joint target list creation) are performed based on multiple sensors or individual sensors. The signal processing steps are named the same for coherent and non-coherent evaluations, but are performed differently. Preferably, angle estimation for coherent data can be performed based on all sensors, while for non-coherent data, angle estimation can be performed based on individual sensor data. Preferably, velocity estimation for coherent data can be performed conventionally, while for non-coherent data, velocity estimation can be additionally complemented by vector velocity estimation. This represents an additional evaluation result. The joint result can be output, for example, in the form of a target list.

[0029] According to a further aspect, the sensor data can be preprocessed in the sensor before being transmitted. For this purpose, a preprocessing unit can be provided in the sensor or the evaluation unit described above can be used in the sensor. The preprocessed sensor data is then transmitted and segmented and evaluated as described above.

[0030] According to a further aspect, the sensor data can be transmitted to an evaluation unit. The evaluation unit can be integrated into the sensor or configured in a central computing device. Preprocessing is then performed in the evaluation unit. Preferably, the sensor data is preprocessed in the evaluation unit, and the evaluation is also performed in these evaluation units. Alternatively, the preprocessed sensor data can be transmitted and split as described above.

[0031] In one form of preprocessing, distance and / or Doppler estimation is performed in each sensor or evaluation unit, for example, by a fast Fourier transform (FFT) or a discrete Fourier transform (DFT). Furthermore, an adapted evaluation with a constant false alarm probability (CFAR) is performed in each sensor using these data, thereby acquiring, i.e., detecting, the target. This adapted CFAR uses a higher false alarm probability compared to conventional methods. A higher false alarm probability results in relatively more targets being detected. A higher false alarm probability should be selected to ensure that as many targets as possible are detected, thereby preventing information about targets in the scene captured by the sensor from being discarded. The preprocessed sensor data is then transmitted. This reduces the data rate during transmission of the preprocessed data by a factor of several, for example, ten times, compared to transmission of raw data.

[0032] In the joint evaluation, an evaluation with a lower constant false alarm probability is advantageously performed based on all sensors. Here, the term "lower" is used in comparison with the CFAR described above. Preferably, a normal parameterization is used for this CFAR, thereby capturing common targets in a normal manner. As described above, based on multiple sensors, non-coherent integration (NCI) can be additionally performed within the spectral range in which the data is transmitted by the multiple sensors. This achieves a significantly higher integration gain compared to individual evaluation. Finally, as described above, further signal processing steps typical for each data type, such as angle estimation and / or velocity calculation, as well as non-coherent processing steps (e.g., vector velocity calculation and / or joint target list creation) are performed based on multiple sensors or individual sensors.

[0033] During preprocessing, each sensor or evaluation unit independently determines which data to discard. Therefore, some sensors may not transmit sensor data. However, angle determination based on missing data is disadvantageous for the overall evaluation of the coherent cooperative radar network. Preferably, a quality criterion is introduced to determine whether targets acquired by only some radar sensors are evaluated. Such a quality criterion may be determined, for example, by a majority vote of the sensors. If the quality criterion is met, the partially acquired targets are evaluated. In this case, it may be possible to perform angle evaluation using only some of the radar sensors. To evaluate non-coherent data, target aggregation may be performed.

[0034] In a further form of sensor data preprocessing, separation and Doppler estimation is performed within each sensor or evaluation unit. Then, a full evaluation with a constant false alarm probability is performed within each sensor or evaluation unit. Preferably, a normal parameterization is used for this CFAR, allowing common targets to be captured in a normal manner. Targets can then be discarded by comparing them with the raw data and the evaluation with a lower CFAR. A peak search can also be performed, and, if necessary, neighboring regions of the peaks can be additionally transmitted, preprocessed, and / or evaluated. Angle estimation is further performed within each sensor or evaluation unit based on the sensor data of the individual sensors and / or bistatic measurement paths. This results in a roughly estimated angle. The results of this preprocessing are stored in a target list, which additionally contains phase information. This additional information relates to the relative phase positions of the individual transmitter-receiver antenna channels for each target and is present as complex amplitudes of all virtual channels. The target list including the phase information is transmitted as the preprocessed sensor data.

[0035] During joint evaluation, the target lists of the individual sensors or evaluation units can be combined. Also, CFAR and / or peak value search can be performed again to reduce the amount of data. If the preprocessed sensor data delivers the same target (e.g., the same separation distance and Doppler data, and possibly additionally the same angle or angle range), cooperative angle estimation can be performed based on the target list. As a result, the computational effort during joint evaluation, and in the best case, during evaluation of coherent data, is significantly reduced. This reduces the data rate during transmission of preprocessed data by a factor of several, e.g., 1000, compared to transmission of raw data. When evaluating individual sensors followed by joint evaluation, the cooperative integration gain before CFAR is relatively small. Finally, as described above, further signal processing steps typical for each data type are performed based on the individual sensors or multiple sensors.

[0036] The coarsely estimated angles obtained in the preprocessing described above can be refined in the joint evaluation because the relative phase positions of the transmit and receive antenna channels are known for all target lists. To this end, coherent cooperative angle estimation is performed with higher angular resolution within a definable range around at least one estimated angle. This can be done, for example, using a fast Fourier transform, a Bartlett estimator, or a maximum likelihood method. Because the angular range over which the higher angular resolution is applied is strongly limited by the separation capabilities of the individual sensors, e.g., in elevation and azimuth angles, the computational effort for this evaluation is comparable to that for conventional angle estimation by individual sensors across the entire angular range. This evaluation also allows, among other things, the detection and separation of multiple targets within the angular range. Typically, the results of this angle estimation can be used instead of the coarse angle estimates of the sensors or evaluation units. If the quality value of the joint evaluation is below the quality value, the angle estimates of the individual sensors or evaluation units can be utilized. The quality value can be obtained, for example, as a result of correlating the control matrix with the complex amplitude. For this purpose, meta-information regarding the processing method can be added to the target list.

[0037] The computer program is specifically designed to perform the steps of the method when executed on a computing device, thereby enabling the method to be implemented in conventional computing devices without the need for structural changes thereto. In addition, the computer program is stored in a machine-readable storage medium.

[0038] A coherent cooperative radar sensor network is further proposed, comprising a plurality of radar sensors, at least two of which operate coherently. The radar sensor network further comprises a data distributor unit capable of distributing sensor data. The data distributor unit may be part of one or more of the sensors. The radar network is designed to perform the steps of the method described above.

[0039] The coherent cooperative radar sensor network may further comprise a central computing device, which is designed to perform the steps of the above-mentioned method, and for this purpose may comprise a data distributor unit.

[0040] Exemplary embodiments of the invention are illustrated in the drawings and explained in more detail in the following description. [Brief explanation of the drawings]

[0041] [Figure 1] FIG. 1 is a schematic diagram of data and data division. [Figure 2] FIG. 1 is a system diagram of a coherent cooperative radar sensor network according to one embodiment of the present invention. [Figure 3] FIG. 1 is a system diagram of a coherent cooperative radar sensor network according to a further embodiment of the present invention. [Figure 4] FIG. 10 is an angle diagram of azimuth and elevation angles for target angle estimation. DETAILED DESCRIPTION OF THE INVENTION

[0042] FIG. 1 shows the basic idea of ​​the method according to the present invention in a data diagram. FIGS. 2 and 3 each show a coherent cooperative radar sensor network comprising multiple radar sensors 11, 12, and 13 (three of which are shown). In this example, the first radar sensor 11 is configured as a non-coherent sensor, while the second and third radar sensors 12 and 13 are configured coherently. In other exemplary embodiments, all radar sensors 11, 12, and 13 may be configured coherently. The sensor data SD1, SD2, and SD3 acquired by the radar sensors 11, 12, and 13 are divided by the data distribution unit 2 according to their evaluation type, i.e., whether they are evaluated coherently or non-coherently. A (two-dimensional) spectrum can be calculated in advance, and the division can be based on this spectrum. Here, the ranges for coherent and non-coherent evaluation are advantageously calculated. For frequency-modulated continuous-wave radar (FMCW), the division can be performed by filtering. Coherent data KD are obtained by high-pass filtering, and non-coherent data NKD are obtained by low-pass filtering. The coherent data KD and non-coherent data NKD are distributed to various evaluation units 110, 120, 130; 31, 32, 33 and evaluated there, see FIGS. 2 and 3. From the non-coherent NKD data, for example, vector velocities are calculated and / or a joint target list is created. With respect to the coherent data KD, phase and / or frequency synchronization is performed, for example via an antenna array, and angle estimation is also performed.

[0043] The data can additionally be divided according to the range, Doppler / velocity and / or angle range of the radar signal used and distributed to evaluation units 110, 120, 130, 31, 32, 33, which then perform the evaluation of the data.

[0044] 2, the radar sensors 11, 12, and 13 each include an evaluation unit 110, 120, and 130, which performs preprocessing of the sensor data SD1, SD2, and SD3. Alternatively, raw data and / or Fourier transformed data, such as range FFT data, Doppler FFT data, and / or range-Doppler FFT data, can be transmitted, and the corresponding steps can be performed later during evaluation. The preprocessed sensor data SD1, SD2, and SD3 include target lists ZL1, ZL2, and ZL3 (see FIG. 4) containing the respective targets Z1, Z2, and Z3. The preprocessed sensor data SD1, SD2, and SD3 (or the raw data and / or Fourier transformed data) are transmitted to a data distributor unit 2. The data distributor unit 2 is configured here as an independent unit, but may also be part of the radar sensors 11, 12, and 13. The data distributor unit 2 divides the data as described above in connection with FIG. 1 and transmits the data to the evaluation units 120, 130, 140 of the radar sensors 11, 12, 13. A bidirectional connection exists between the radar sensors 11, 12, 13 and the data distributor unit 2. Here, the non-coherent data NKD is transmitted to the evaluation unit 110 of the first radar sensor 11. This applies, in particular, when the data distributor unit 2 is part of the radar sensors 11, 12, 13. The coherent data KD is separated according to the selected target range or subspectrum, divided accordingly, and transmitted to the evaluation unit 120 of the second radar sensor 12 and the evaluation unit 130 of the third radar sensor 13. These radar sensors 12, 13 operate coherently. The evaluation units 110, 120, 130 then evaluate the data as described above in connection with FIG. 1. The evaluated data is then transmitted again to the central computing device 3 via the data distributor unit 2. The central computing device 3 comprises an evaluation unit 30 which generates a joint evaluation result. Alternatively, this can also be done in one of the evaluation units 110, 120, 130 of the radar sensors 11, 12, 13.

[0045] In FIG. 3 , the radar sensors 11, 12, 13 each include a pre-processing unit 111, 121, 131 that performs pre-processing of the sensor data SD1, SD2, SD3. Alternatively, raw data and / or Fourier transformed data, such as range FFT data, Doppler FFT data, and / or range-Doppler FFT data, can be transmitted, and the corresponding steps can be performed later during evaluation. The pre-processed sensor data SD1, SD2, SD3 include target lists ZL1, ZL2, ZL3 (see FIG. 4 ) that include the respective targets Z1, Z2, Z3. The pre-processed sensor data SD1, SD2, SD3 (or the raw data and / or Fourier transformed data) are transmitted to the data distributor unit 2. A central computing device 3 comprising multiple evaluation units 30, 31, 32, 33 is also contemplated. The evaluation units 30, 31, 32, 33 may be, for example, processors, virtual cores, or software blocks. The data distributor unit 2 is configured as an independent unit here, but may also be part of the central computing device 3. The data distributor unit 2 divides the data as described above in connection with FIG. 1 and transmits the data to evaluation units 31, 32, and 33 of the central computing device 3. The connections between the radar sensors 11, 12, and 13 and the data distributor unit 2, as well as the connections between the data distributor unit 2 and the central computing device 3, may be configured as simple connections. Purely by way of example, the non-coherent data NKD is transmitted to the evaluation unit 31. The coherent data KD is separated according to selected target or partial spectrum ranges, divided accordingly, and transmitted to the evaluation units 32 and 33. The evaluation units 31, 32, and 33 then evaluate the data as described above in connection with FIG. 1. The evaluated data is then transmitted to the evaluation unit 30, which performs a combined evaluation, or overall evaluation.

[0046] Figure 4 shows three target lists ZL1, ZL2, and ZL3 for three radar sensors 11, 12, and 13. In this example, each list contains three targets Z1, Z2, and Z3. At least the first target Z1 should be the same in the three target lists ZL1, ZL2, and ZL3 for the three radar sensors 11, 12, and 13. An angle diagram of the azimuth angle φ and the elevation angle θ is shown. During preprocessing, the first radar sensor 11 estimated the azimuth angle φ1 and the elevation angle θ1 for the first target Z1 using coarse angle estimation with low resolution. In the overall evaluation, coherent cooperative angle estimation is performed for all radar sensors 11, 12, and 13. Since the coarse angles φ1 and θ1 are already known, angle estimation is performed with higher angular resolution within a definable range around them. The higher angular resolution is represented by a finer grid in Figure 4. The range limits are selected for azimuth angle φ at φ1-Δφ and φ1+Δφ, and for elevation angle θ at θ1-Δθ and θ1+Δθ.

Claims

1. A method for controlling a coherent cooperative radar sensor network comprising a plurality of radar sensors (11, 12, 13), at least two of which operate coherently, comprising:

1. A method according to claim 1, wherein the sensor data (SD1, SD2, SD3) are divided into data to be evaluated coherently (KD) and data to be evaluated non-coherently (NKD) depending on the evaluation type, and the data to be evaluated coherently (KD) and the data to be evaluated non-coherently (NKD) are transmitted to different evaluation units (110, 120, 130; 31, 32, 33), which then perform their own evaluations, and the individual evaluations are combined to generate an overall evaluation.

2. 2. The method according to claim 1, characterized in that the sensor data (SD1, SD2, SD3) are divided according to the range range, Doppler / velocity range and / or angle range of the radar signal used, and the data of different ranges are transmitted to different evaluation units (110, 120, 130; 31, 32, 33), which then each perform an evaluation, and the individual evaluations are combined to generate an overall evaluation.

3. 3. The method according to claim 1, wherein the sensor data (SD1, SD2, SD3) are transmitted to a data distributor unit (2), the sensor data (SD1, SD2, SD3) are divided by the data distributor unit (2), and the data to be evaluated coherently (KD) and the data to be evaluated non-coherently (NKD) are distributed by the data distributor unit (2) to the evaluation units (110, 120, 130; 31, 32, 33).

4. 4. The method according to claim 1, wherein the data (KD, NKD) are distributed to evaluation units (110, 120, 130) located at the sensors (11, 12, 13), and the evaluation is carried out by the evaluation units (110, 120, 130) of the sensors (11, 12, 13).

5. 5. The method according to claim 1, wherein the data (KD, NKD) are transmitted to a central computing device (3), and an evaluation unit (31, 32, 33) of the central computing device (3) performs the evaluation.

6. 6. The method according to claim 1, wherein the sensor data (SD1, SD2, SD3) are transmitted from the sensors (11, 12, 13) as raw data or as Fourier transformed data.

7. 6. The method according to claim 1, wherein the sensor data (SD1, SD2, SD3) are preprocessed in the sensors (11, 12, 13) and the sensor data (SD1, SD2, SD3) are transmitted as preprocessed sensor data.

8. 7. The method according to claim 1, wherein the sensor data (SD1, SD2, SD3) are pre-processed in the evaluation unit (110, 120, 130).

9. 9. The method according to claim 7 or 8, characterized in that during the pre-processing of the sensor data (SD1, SD2, SD3) a separation distance and Doppler estimation is carried out, from which an adapted estimation leads to acquisition of the targets (Z1, Z2, Z3) with a constant probability of false alarm.

10. 10. The method of claim 9, wherein during the overall evaluation, a further evaluation is performed with a lower constant false alarm probability in order to capture a common target (Z1).

11. 11. The method according to claim 9 or 10, characterized in that targets acquired by only some of the radar sensors (11, 12, 13) are evaluated only if a quality criterion is met.

12. 9. The method according to claim 7 or 8, characterized in that during the pre-processing of the sensor data (SD1, SD2, SD3), a separation and Doppler estimation is carried out, from which target acquisition and angle estimation are performed by complete evaluation with a constant false alarm probability, and the transmitted sensor data comprises a target list (ZL1, ZL2) including phase information.

13. At least one angle (φ １ , θ １ 13. The method of claim 12, wherein during the evaluation of the at least one estimated angle, a coherent cooperative angle estimation is performed at a higher angular resolution within a definable range around the at least one estimated angle.

14. A computer program designed to carry out the steps of the method according to any one of claims 1 to 13.

15. A machine-readable storage medium having stored thereon the computer program of claim 14.

16. 14. A coherent cooperative radar sensor network comprising a plurality of radar sensors (11, 12, 13) in which at least two sensors (12, 13) operate coherently, and a data distributor unit (2), the radar network being designed to perform the method according to any one of claims 1 to 13.

17. A coherent cooperative radar sensor network according to claim 16, characterized by a central computing device (3) designed to carry out the method according to any one of claims 1 to 13.

Images