Millimeter wave radar signal processing method, storage medium, radar equipment and vehicle

Abstract

The present application discloses a millimeter wave radar signal processing method, storage medium, radar equipment and vehicle, and relates to the field of radar signal processing technology. The method is applied to a hardware accelerator, including: obtaining one-dimensional Fourier data corresponding to each detection target; when a digital signal processor extracts amplitude and phase values ​​from two-dimensional Fourier data corresponding to a previous detection target, performing a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target; obtaining two-dimensional Fourier data corresponding to the current detection target; returning to the step of performing two-dimensional Fourier transform until the digital signal processor obtains amplitude and phase values ​​corresponding to all detection targets. The present application uses a hardware accelerator and a digital signal processor to perform two-dimensional Fourier transform calculations and amplitude and phase value extraction on each detection target in an interactive and parallel manner, thereby improving signal processing efficiency and avoiding the problem of low memory utilization caused by opening up memory space according to the maximum number of detection targets.

Description

Technical Field

[0001] The present application relates to the field of radar signal processing technology, and in particular to a millimeter wave radar signal processing method, a storage medium, a radar device and a vehicle. Background Art

[0002] With the rapid development of intelligent driving technology, compared with cameras and lidar, millimeter-wave radar occupies an important position in advanced driver assistance systems due to its relatively low cost and excellent robustness in harsh environments.

[0003] In the millimeter-wave radar used as a vehicle-mounted radar, after receiving the original analog-to-digital converter (ADC) sampling signal data, the RF front end can obtain the corresponding estimation information by performing Fast Fourier Transformation (FFT) processing on the three-dimensional information of the distance, speed and angle of the detected target. In this signal processing process, usually, FFT processing is performed on any two of the three dimensions (distance and speed) to obtain the spectrum heat map of the corresponding dimensions, and then target detection is performed based on the spectrum heat map. After obtaining the detected target, FFT processing of two dimensions is performed again, so as to obtain the information of the third dimension (angle) in combination with the angle measurement algorithm.

[0004] In engineering implementation, a hardware accelerator (HWA) can be integrated on the radar chip to cooperate with the digital signal processor (DSP) integrated on the radar chip to implement the above signal processing process. However, since the existing HWA and DSP are usually designed as serial links, after obtaining the detection target in the signal processing process, it is necessary to open up memory space according to the maximum number of detection targets and perform two-dimensional FFT processing on each detection target, resulting in low memory utilization and slow signal processing speed. Summary of the invention

[0005] The main purpose of the present application is to provide a millimeter wave radar signal processing method, storage medium, radar equipment and vehicle, aiming to solve the technical problems of low memory utilization and slow signal processing speed in the existing radar signal processing process based on HWA and DSP serial link implementation, due to the need to open up memory space according to the maximum number of detected targets and perform FFT processing on each detected target.

[0006] To achieve the above objectives, the present application proposes a millimeter wave radar signal processing method, which is applied to a hardware accelerator integrated on a radar chip. The method includes:

[0007] Obtaining one-dimensional Fourier data corresponding to each detection target, wherein the one-dimensional Fourier data is obtained by converting the original received signal;

[0008] When the digital signal processor extracts the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the last detected target, a two-dimensional Fourier transform is performed on the one-dimensional Fourier data corresponding to the current detected target;

[0009] Obtaining two-dimensional Fourier data corresponding to the current detection target, and sending the two-dimensional Fourier data corresponding to the current detection target to the digital signal processor;

[0010] Return to the step of performing a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target when the digital signal processor extracts the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the previous detection target, until the digital signal processor obtains the amplitude and phase values ​​corresponding to all detection targets.

[0011] In one embodiment, the step of obtaining one-dimensional Fourier data corresponding to each detection target includes:

[0012] Performing fast Fourier transform on the original received signal along the range dimension and the Doppler dimension respectively to obtain a range-Doppler spectrum diagram;

[0013] Based on the range-Doppler spectrum diagram, a constant false alarm detection algorithm is used to perform target detection to obtain constant false alarm detection information;

[0014] A plurality of detection targets are obtained according to the constant false alarm detection information, and one-dimensional Fourier data corresponding to each of the detection targets is determined.

[0015] In one embodiment, the constant false alarm detection information includes: a distance index of each of the detection targets;

[0016] The step of performing a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target includes:

[0017] Perform a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target according to the distance index of the current detection target.

[0018] In one embodiment, after the step of performing one-dimensional Fourier transform on the original received signal to obtain one-dimensional Fourier data corresponding to each of the detection targets, the method further includes:

[0019] The range-Doppler spectrum diagram is stored in a memory unit, so that the digital signal processor obtains the range-Doppler spectrum diagram and the constant false alarm detection information through the memory unit.

[0020] In addition, to achieve the above purpose, the present application also proposes a millimeter wave radar signal processing method, which is applied to a digital signal processor integrated on a radar chip, and the method includes:

[0021] When the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the next detection target, receiving the two-dimensional Fourier data corresponding to the current detection target;

[0022] Extracting amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase value corresponding to the current detection target;

[0023] Return to the step of receiving the two-dimensional Fourier data corresponding to the current detection target when the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the next detection target, until the amplitude and phase values ​​corresponding to all the detection targets are obtained.

[0024] In one embodiment, before the step of extracting amplitude and phase values ​​from the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target, the step includes:

[0025] Acquire a range-Doppler spectrum diagram and constant false alarm detection information, wherein the constant false alarm detection information includes: a Doppler index of each detected target;

[0026] Correspondingly, the step of extracting the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target includes:

[0027] The range-Doppler spectrum diagram corresponding to the current detection target is demodulated according to the code type information, and the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target are extracted according to the Doppler index of the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target.

[0028] In one embodiment, the constant false alarm detection information further includes: a distance index of each of the detection targets, and after the step of obtaining the amplitude and phase values ​​corresponding to all the detection targets, the following steps are included:

[0029] According to the distance index and Doppler index of each detection target, each detection target is screened by using a preset strategy to obtain a screened detection target;

[0030] A preset angle measurement algorithm is used to perform angle measurement calculation on the amplitude and phase values ​​corresponding to each of the screened detection targets to obtain angle measurement results corresponding to each of the screened detection targets.

[0031] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, on which a millimeter-wave radar signal processing program is stored, and when the millimeter-wave radar signal processing program is run, the steps of the millimeter-wave radar signal processing method described above are implemented.

[0032] In addition, to achieve the above-mentioned purpose, the present application also proposes a radar device, which includes: a radar body, a hardware accelerator and a digital signal processor, wherein the hardware accelerator is used to execute the millimeter-wave radar signal processing method applied to the hardware accelerator integrated on the radar chip as described above, and the digital signal processor is used to execute the millimeter-wave radar signal processing method applied to the digital signal processor integrated on the radar chip as described above.

[0033] In addition, to achieve the above-mentioned purpose, the present application also proposes a vehicle, which includes: a vehicle body, and a radar device as described above installed on the vehicle body.

[0034] The millimeter wave radar signal processing method disclosed in the present application is applied to a hardware accelerator integrated on a radar chip, and includes: obtaining one-dimensional Fourier data corresponding to each detection target, wherein the one-dimensional Fourier data is converted from an original received signal; when a digital signal processor extracts amplitude and phase values ​​of two-dimensional Fourier data corresponding to a previous detection target, performing two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to a current detection target; obtaining two-dimensional Fourier data corresponding to the current detection target, and sending the two-dimensional Fourier data corresponding to the current detection target to the digital signal processor; returning to perform the step of performing two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target when the digital signal processor extracts amplitude and phase values ​​of two-dimensional Fourier data corresponding to a previous detection target, until the digital signal processor obtains amplitude and phase values ​​corresponding to all detection targets. Because in the method of the present application, the hardware accelerator can perform a two-dimensional Fourier transform on the one-dimensional Fourier data of the current detection target while the digital signal processor extracts the amplitude and phase values ​​of the two-dimensional Fourier data of the previous detection target, thereby obtaining the two-dimensional Fourier data of the current detection target, and thus the two-dimensional Fourier transform calculation can be implemented for each detection target in an interactive and parallel manner using the hardware accelerator and the digital signal processor, effectively reducing the processing time when processing the same number of detection targets, and avoiding the problem of low memory utilization caused by opening up memory space according to the maximum number of detection targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0037] Figure 1 This is a flow chart of a first embodiment of a method for processing millimeter-wave radar signals in a hardware accelerator integrated on a radar chip according to the present application;

[0038] Figure 2 A schematic diagram showing a comparison between the parallel signal processing link of the present application and the existing serial signal processing link;

[0039] Figure 3 This is a flow chart of a first embodiment of a method for processing millimeter wave radar signals in a digital signal processor integrated on a radar chip according to the present application;

[0040] Figure 4 This is a flow chart of a second embodiment of a method for processing millimeter-wave radar signals in a digital signal processor integrated on a radar chip according to the present application;

[0041] Figure 5 This is a schematic diagram of the entire process of the millimeter wave radar signal processing method of this application;

[0042] Figure 6 This is the spatial spectrum of each detected target in the experimental example of the millimeter wave radar signal processing method of this application.

[0043] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0044] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0045] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0046] The present application embodiment provides a millimeter wave radar signal processing method, which is applied to a hardware accelerator integrated on a radar chip, referring to Figure 1 , Figure 1 This is a flow chart of a first embodiment of a method for processing millimeter-wave radar signals in a hardware accelerator integrated on a radar chip according to the present application.

[0047] In this embodiment, the method includes: steps S10 to S40:

[0048] Step S10: obtaining one-dimensional Fourier data corresponding to each detection target, wherein the one-dimensional Fourier data is obtained by converting the original received signal.

[0049] It should be noted that the execution subject of this embodiment may be a hardware accelerator (HWA) integrated on a radar chip, and a digital signal processor (DSP) may also be integrated on the radar chip. The HWA may implement functions such as FFT, target detection, compression, and histogram statistics, and the DSP may implement functions such as signal source number estimation and super-resolution angle measurement algorithm, and the HWA and DSP may jointly implement functions such as FFT angle measurement and velocity ambiguity resolution.

[0050] It should be understood that the original received signal may be an original ADC sampling signal received by the hardware accelerator (HWA), and may include a frame of original ADC sampling signals with a certain number of chirps as a signal processing unit.

[0051] It should also be noted that, since the multiple-input multiple-output (MIMO) technology of multiple transmitting antennas and multiple receiving antennas has the advantage of improving target detection performance and target parameter estimation performance compared to the single-input single-output technology, and for the transmitting end with multiple antennas, the Doppler diversity multiple access (DDMA) waveform has the advantages of good orthogonality and low side lobes in the distance / speed dimension. Therefore, the original ADC sampling signal in the embodiment of the present application can be obtained based on the DDMA waveform scheme.

[0052] It is understandable that after obtaining the original ADC sampling signal, HWA can perform FFT on the original ADC sampling signal along the fast time distance dimension and the slow time Doppler dimension (speed dimension) to obtain a range-Doppler spectrum map (RVmap). Then, based on the range-Doppler spectrum map, a constant false alarm rate (CFAR) algorithm is used to perform target detection to obtain constant false alarm detection information (CFAR information), and then the number of detected targets, target distance information, and target Doppler information can be determined from the CFAR information.

[0053] The CFAR algorithm may be, for example, a CA-CFAR algorithm or an OS-CFAR algorithm, which is not limited in this embodiment.

[0054] The one-dimensional Fourier data (1DFFT data), RVmap and CFAR information obtained by performing FFT on the original ADC sampling signal along the distance dimension of fast time can be stored in the memory unit (such as L3 memory) of the radar chip. The RVmap is obtained by performing two-dimensional Fourier transform (2DFFT) on the 1DFFT data and then extracting the amplitude. This is because the two-dimensional Fourier data (2DFFT data) obtained by performing 2DFFT on the 1DFFT data involves the phase value information of each detection target, and its data volume is large and therefore difficult to be stored in the L3 memory.

[0055] Therefore, for each detection target, HWA can directly extract the 1DFFT data corresponding to the detection target from the L3 memory, so as to implement the 2DFFT of the 1DFFT data again through the CFAR information. And DSP can also obtain RVmap from the L3 memory to demodulate it.

[0056] Step S20: When the digital signal processor extracts amplitude and phase values ​​from the two-dimensional Fourier data corresponding to the last detected target, a two-dimensional Fourier transform is performed on the one-dimensional Fourier data corresponding to the current detected target.

[0057] It should be noted that each detection target can be regarded as an independent processing unit (Block), the current detection target can correspond to the nth Block (n∈N, N is the total number of detection targets), and the previous detection target of the current detection target can correspond to the n-1th Block. Each Block may include 1DFFT data corresponding to each detection target, target distance information and target Doppler information, the target distance information may be a distance index (rngIdx), and the target Doppler information may be a Doppler index (dopIdx).

[0058] It is understandable that if the current detection target is the first Block, that is, there is no previous detection target of the current detection target, the HWA can directly perform 2DFFT on the first Block, and the DSP does not perform any operation at this time.

[0059] If the current detection target is not the first block, the HWA needs to wait for the n-1th block to complete the 2DFFT before performing the 2DFFT on the nth block. When the HWA performs the 2DFFT on the nth block, the DSP can extract the amplitude and phase values ​​of the n-1th block.

[0060] In a specific implementation, when the DSP extracts the amplitude and phase values ​​of the 2DFFT data corresponding to the last detected target, the HWA may perform 2DFFT on the 1DFFT data corresponding to the current detected target according to the rngIdx of the current detected target.

[0061] Step S30: obtaining two-dimensional Fourier data corresponding to the current detection target, and sending the two-dimensional Fourier data corresponding to the current detection target to the digital signal processor.

[0062] It can be understood that HWA performs 2DFFT on the 1DFFT data corresponding to the current detection target according to the rngIdx of the current detection target, thereby obtaining the 2DFFT data corresponding to the current detection target, and can send the 2DFFT data to the DSP so that the DSP continues to extract the amplitude and phase values ​​of the 2DFFT data.

[0063] It should also be noted that, compared with the existing FFT processing process of multiple dimensions directly performed in the DSP, this embodiment uses the advantage of HWA for FFT batch processing, so that the DSP and HWA can perform interactive processing, thereby improving signal processing efficiency.

[0064] Step S40: Return to the step of performing two-dimensional Fourier transform on one-dimensional Fourier data corresponding to the current detection target when the digital signal processor extracts amplitude and phase values ​​on two-dimensional Fourier data corresponding to the previous detection target, until the digital signal processor obtains amplitude and phase values ​​corresponding to all detection targets.

[0065] It can be understood that HWA repeats the above steps S20 to S30 until 2DFFT is completed for the 1DFFT data corresponding to all detection targets to obtain the 2DFFT data corresponding to each detection target, and DSP completes the amplitude and phase value extraction of the 2DFFT data corresponding to each detection target, thereby achieving the interactive and parallel completion of 2DFFT and amplitude and phase value extraction of N blocks by HWA and DSP. At this time, the amplitude and phase values ​​corresponding to each detection target are stored in DSP, and the angle measurement algorithm is applied to the amplitude and phase values ​​in DSP, so that the target angle information corresponding to each detection target can be obtained.

[0066] Further, please refer to Figure 2 , the HWA and DSP interactive parallel signal processing link of the present application is compared with the existing HWA and DSP serial signal processing link, Figure 2 This is a schematic diagram comparing the parallel signal processing link of the present application with the existing serial signal processing link.

[0067] exist Figure 2In the figure, 2-a is the existing HWA and DSP serial signal processing link, and 2-b is the HWA and DSP parallel signal processing link in this application. The white block (Block) indicates that HWA processes the Block, that is, 2DFFT calculation is performed by HWA; the gray block (Block) indicates that DSP processes the Block, that is, DSP extracts the amplitude and phase values ​​of 2DFFT data. Take the total number of detected targets as 3 as an example.

[0068] In 2-a, after obtaining the 1DFFT data corresponding to each detection target, HWA first processes the n-1th Block; after HWA processes the n-1th Block, DSP processes the n-1th Block. After DSP processes the n-1th Block, HWA processes the nth Block; after HWA processes the nth Block, DSP processes the nth Block. After DSP processes the n-1th Block, HWA processes the n+1th Block; after HWA processes the n+1th Block, DSP processes the n+1th Block.

[0069] In the above serial signal processing link, HWA and DSP process each Block in turn: when the current Block is processed, the next Block will be processed.

[0070] In 2-b, after obtaining the 1DFFT data corresponding to each detection target, HWA first processes n-1 blocks. After HWA processes the n-1th block, HWA directly processes the nth block; at the same time, DSP processes the n-1th block. After HWA processes the nth block, HWA directly processes the n+1th block; at the same time, DSP processes the nth block. After HWA processes the n+1th block, DSP processes the n+1th block.

[0071] In the above parallel signal processing link, HWA and DSP interact to process each Block: HWA and DSP can process different Blocks at the same time.

[0072] From the comparison between 2-a and 2-b, it can be seen that compared with the existing HWA and DSP serial signal processing link, the HWA and DSP parallel signal processing link adopted in this application fully utilizes the advantage of HWA for FFT batch processing, reduces the processing time when processing the same amount of data, and the required processing time is greatly reduced with the number of Blocks (detection targets), avoiding the operation of opening up memory space according to the maximum number of detection targets.

[0073] In this embodiment, the hardware accelerator can perform a two-dimensional Fourier transform on the one-dimensional Fourier data of the current detection target while the digital signal processor extracts the amplitude and phase values ​​of the two-dimensional Fourier data of the previous detection target, so as to obtain the two-dimensional Fourier data of the current detection target, so that the two-dimensional Fourier transform calculation and amplitude and phase value extraction can be realized for each detection target in an interactive and parallel manner between the hardware accelerator and the digital signal processor. Compared with the existing hardware accelerator and digital signal processor serial signal processing link, the processing time is effectively reduced when processing the same number of detection targets, and the problem of low memory utilization caused by opening up memory space according to the maximum number of detection targets is avoided.

[0074] The present application also provides a millimeter wave radar signal processing method, which is applied to a digital signal processor integrated on a radar chip, referring to Figure 3 , Figure 3 This is a flow chart of a first embodiment of a method for processing millimeter-wave radar signals in a digital signal processor integrated on a radar chip according to the present application.

[0075] In this embodiment, the method includes: steps A10 to A30:

[0076] Step A10: When the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the next detection target, the two-dimensional Fourier data corresponding to the current detection target is received.

[0077] It should be noted that the execution subject of this embodiment may be a digital signal processor (DSP) integrated on the radar chip. The radar chip may also be integrated with a hardware accelerator (HWA). The HWA may implement functions such as FFT, target detection, compression, and histogram statistics, and the DSP may implement functions such as signal source number estimation and super-resolution angle measurement algorithm, and the HWA and DSP may also jointly implement functions such as FFT angle measurement and velocity ambiguity resolution.

[0078] It should be understood that each detection target can be regarded as an independent processing unit (Block), the current detection target can correspond to the nth Block, and the next detection target of the current detection target can correspond to the n+1th Block.

[0079] It can be understood that, while the HWA performs 2DFFT on the (n+1)th Block, the DSP can extract the amplitude and phase values ​​of the nth Block.

[0080] In the specific implementation, after HWA completes 2DFFT on the 1DFFT data corresponding to the current detection target, it then performs 2DFFT on the 1DFFT data corresponding to the next detection target. At the same time, DSP can obtain the 2DFFT data corresponding to the current detection target sent by HWA for subsequent amplitude and phase value extraction.

[0081] Step A20: extracting amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target.

[0082] It should be noted that before extracting the amplitude and phase values, the DSP can also obtain the range-Doppler spectrum map (RVmap) and constant false alarm detection information (CFAR information) from the L3 memory of the radar chip. The CFAR information includes the range index (rngIdx) and the Doppler index (dopIdx).

[0083] It should also be noted that, since the original ADC sampling signals corresponding to each detection target in this embodiment are obtained based on the DDMA waveform scheme, the DDMA waveform scheme adopts a mode in which multiple transmitting antennas transmit simultaneously and multiple receiving antennas receive simultaneously. Therefore, before applying the angle measurement algorithm to the amplitude and phase values ​​to obtain the target angle information, it is necessary to identify the transmission channel number to obtain the amplitude and phase values ​​corresponding to each channel of the virtual array. Therefore, step A20 specifically includes:

[0084] Step A200: demodulate the range-Doppler spectrum diagram corresponding to the current detection target according to the code type information, and extract the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target according to the Doppler index of the current detection target to obtain the amplitude and phase value corresponding to the current detection target.

[0085] It should be noted that the code type information may be a unique corresponding code sequence designed in the DDMA waveform scheme and allocated to each transmit / receive channel. These code sequences are orthogonal or quasi-orthogonal to ensure that the signals can be correctly separated and identified.

[0086] It can be understood that since RVmap is obtained by performing FFT on the original received signal obtained by each receiving antenna based on the distance dimension and the Doppler dimension, and different detection targets can be obtained by signals obtained by different receiving antennas, the RVmap corresponding to the current detection target can be demodulated according to the code type information, and then the 2DFFT data corresponding to the current detection target can be extracted according to dopIdx to ensure that the obtained amplitude and phase values ​​contain channel information (that is, the amplitude and phase values ​​are obtained by rearranging the channels according to the demodulation information), so as to accurately obtain the angle measurement calculation results corresponding to each detection target.

[0087] It should also be noted that, since specific code type information is used for demodulation in this embodiment, compared with the traditional method of adding empty bands on the RVmap for demodulation, there is no need to sacrifice Doppler dimension information, thereby largely avoiding the influence of spectrum aliasing on angle measurement accuracy when there are many points. The method of this embodiment is not limited to a specific DDMA waveform scheme and is universal.

[0088] In addition, the method of this embodiment can also be used for other waveform schemes, such as the time division multiple access (TDMA) waveform scheme: when processing the signal under the TDMA waveform scheme, the code information can be replaced with the time division information, and the RVmap can be demodulated based on the time division information, and then the amplitude and phase values ​​can be extracted.

[0089] Step A30: Return to the step of receiving the two-dimensional Fourier data corresponding to the current detection target when the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the next detection target, until the amplitude and phase values ​​corresponding to all the detection targets are obtained.

[0090] It can be understood that the DSP repeatedly executes the above steps A100 to A200 until the amplitude and phase value extraction of the 2DFFT data corresponding to each detection target is completed, thereby achieving the completion of the 2DFFT and amplitude and phase value extraction of all blocks by the interaction of the HWA and the DSP.

[0091] In this embodiment, the digital signal processor can extract the amplitude and phase values ​​of the two-dimensional Fourier data of the current detection target while the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data of the next detection target, so that the amplitude and phase value extraction can be realized for each detection target in an interactive and parallel manner between the hardware accelerator and the digital signal processor, and because the range-Doppler spectrum diagram is demodulated in advance based on the code type information, it is ensured that the amplitude and phase value contains the corresponding channel information for subsequent angle estimation. Compared with the existing method of reserving empty Doppler sub-bands for decoding in DDMA waveform design, the method of this embodiment avoids the risk of affecting radar performance due to sacrificing the maximum unambiguous speed, is applicable to different DDMA waveform design schemes, and improves the flexibility of early waveform design.

[0092] Based on the above embodiments, a second embodiment of the millimeter wave radar signal processing method of the present application applied to a digital signal processor integrated on a radar chip is proposed, referring to Figure 4 , Figure 4 This is a flow chart of a second embodiment of a method for processing millimeter-wave radar signals in a digital signal processor integrated on a radar chip according to the present application.

[0093] In this embodiment, in order to further improve the signal processing speed, after step A30, the following steps are further included: Steps SA401 to SA402:

[0094] Step A401: According to the distance index and Doppler index of each detection target, a preset strategy is used to screen each detection target to obtain screened detection targets.

[0095] It should be noted that the preset strategy may be a screening strategy set based on different factors, which may include: signal strength threshold, signal-to-noise ratio, confidence level of the detection target, historical detection data, etc.

[0096] In the specific implementation, DSP screens each detection target according to the rngIdx and dopIdx of each detection target and the screening conditions in the preset strategy, removes possible false alarms and unimportant targets, and retains the detection results that are most likely to represent the real target.

[0097] Step A402: using a preset angle measurement algorithm to perform angle measurement calculation on the amplitude and phase values ​​corresponding to each of the screened detection targets, to obtain angle measurement results corresponding to each of the screened detection targets.

[0098] It can be understood that the preset angle measurement algorithm can be, for example, an angle measurement algorithm such as DBF, 3D-FFT, etc., which can be selected on demand according to user needs, and this embodiment does not limit this.

[0099] In the specific implementation, DSP uses the angle measurement algorithm to perform angle measurement calculations on the amplitude and phase values ​​corresponding to each screened detection target, and obtains the angle measurement results corresponding to each screened detection target, which helps to improve the reliability of the target angle information in subsequent data analysis and decision-making for each detection target.

[0100] The digital signal processor of this embodiment uses a preset strategy to screen each of the detection targets according to the distance index and Doppler index of each of the detection targets to obtain the screened detection targets; and uses a preset angle measurement algorithm to perform angle measurement calculation on the amplitude and phase values ​​corresponding to each of the screened detection targets to obtain the angle measurement results corresponding to each of the screened detection targets. Since the amplitude and phase values ​​corresponding to each detection target are obtained and angle measurement calculation is performed in the digital signal processor in this embodiment, the detection targets can be screened by the preset strategy before the angle measurement calculation, which reduces the time consumption of the angle measurement calculation and increases the degree of freedom of the signal processing process.

[0101] Further, reference may also be made to Figure 5 The whole process of the millimeter wave radar signal processing method of this application is described. Figure 5 This is a schematic diagram of the entire process of the millimeter wave radar signal processing method of this application.

[0102] based on Figure 5 For a radar with N transmitting antennas and M receiving antennas, the code information of the N transmitting antennas can be configured in a vector C, where C = (C1, C2, ... C N ). The 1DFFT data of the M receiving antennas can be stored in the L3 memory and the L3 memory area is marked as 1DCompResult, the CFAR information obtained after the constant false alarm detection is stored in the L3 memory area marked as cfarResult, and the RVmap is stored in the L3 memory area marked as cfarResult.

[0103] In HWA, each detection target is determined according to the CFAR information in cfarResult, and the 1D FFT data corresponding to each detection target is determined in combination with 1DCompResult.

[0104] S1. HWA determines the 1DFFT data, DDMA code information, RVmap, and CFAR information corresponding to each detection target as the Block information of each detection target, and synchronizes the Block information to DSP.

[0105] S2, using HWA and DSP ping-pong interactive parallel mode to perform 2DFFT and extract amplitude and phase values ​​on the 1DFFT data corresponding to each detection target. Specifically including:

[0106] S201, HWA performs 2DFFT (ping) on ​​the 1DFFT data in the nth Block to obtain the corresponding 2DFFT data; at the same time, DSP demodulates the RVmap corresponding to the n-1th Block and extracts the amplitude and phase values ​​(pong) of the 2DFFT data corresponding to the n-1th Block;

[0107] S202, HWA performs 2DFFT (pong) on ​​the 1DFFT data in the n+1th Block to obtain the corresponding 2DFFT data; at the same time, DSP demodulates the RVmap corresponding to the nth Block and extracts the amplitude phase value (ping) of the 2DFFT data corresponding to the nth Block;

[0108] S203, HWA performs 2DFFT (ping) on ​​the 1DFFT data in the n+2th Block to obtain the corresponding 2DFFT data; at the same time, DSP demodulates the RVmap corresponding to the n+1th Block and extracts the amplitude and phase values ​​(pong) of the 2DFFT data corresponding to the n+1th Block.

[0109] S3. When the 2DFFT data corresponding to each detection target is obtained in the DSP, the DSP screens the detection target points through a preset strategy, outputs the number of screened detection targets, and performs angle measurement calculation based on the corresponding amplitude and phase values.

[0110] Furthermore, specific experimental data may be used here to verify the feasibility of the millimeter wave radar signal processing method of the present application.

[0111] Experimental conditions: The number of radar transmitting antennas N = 4, the number of receiving antennas M = 4, the DDMA code configuration is (C1 = 0, C2 = 180, C3 = 240, C4 = 330), the number of chirps in one frame of the original ADC signal chirpNum = 512, the number of sampling points sampleNum = 512, and the experiment simulates 3 targets (all valid and qualified targets. CFAR information includes: distance index rngIdx = {5, 7, 9}, and Doppler index dopIdx = {13, 24, 20}. The angles of the three targets are known to be: -53 degrees, 42 degrees and -36 degrees.

[0112] Based on the above Figure 5 As shown in S1, the code information of the four transmitting antennas is stored in the vector C. After processing one frame of radar signal, the 1DFFT data of the four receiving antennas is stored in 1DCompResult, the CFAR information is stored in cfarResult, and the RVmap is stored in rdResult.

[0113] For example, in S2, three Blocks (rngIdx is 5, 7, and 9 respectively) are input in sequence. The first Block (rngIdx is 5) is executed first. Since it is the first Block, there is no need to wait. HWA directly calculates the 2DDFT of this Block (pong) and DSP does not perform any operation. After the second Block (rngIdx is 7) is input and HWA completes the 2DFFT of the first Block (pong), HWA executes the 2DFFT of the second Block (ping). At this time, DSP extracts the amplitude and phase values ​​of the first Block (rngIdx is 5).

[0114] Repeat the above operations, and the HWA and DSP interact to complete the 2DFFT calculation and amplitude and phase information extraction of the three blocks. For example, in S3, the amplitude and phase values ​​of the above three detection targets are stored in the DSP. After the preset strategy screening, it is determined that the above three detection targets are all important targets, and the angle measurement calculation can be performed according to the amplitude and phase values ​​corresponding to each detection target.

[0115] The spatial spectra of the three detection targets obtained in the angle measurement calculation are as follows: Figure 6 As shown, Figure 6 This is the spatial spectrum of each detected target in the experimental example of the millimeter wave radar signal processing method of this application. Figure 6 The X-axis of each coordinate axis can represent the angle, and the Y-axis of each coordinate axis can represent the amplitude value of the signal. Figure 6 The spatial spectrogram corresponding to each detected target is interpolated and angle indexed, and the angle measurement results corresponding to each detected target are obtained as follows: -53.26 degrees, 41.80 degrees, and -36.29 degrees. Since the angles of the three targets are known to be: -53 degrees, 42 degrees, and -36 degrees, the above angle measurement result errors meet the angle measurement accuracy, which verifies that the millimeter wave radar signal processing method of the present application is feasible.

[0116] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the millimeter wave radar signal processing method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0117] The present application also provides a radar device, which includes: a radar body, a hardware accelerator and a digital signal processor, wherein the hardware accelerator is used to execute the millimeter-wave radar signal processing method applied to the hardware accelerator integrated on the radar chip as described in the above embodiment, and the digital signal processor is used to execute the millimeter-wave radar signal processing method applied to the digital signal processor integrated on the radar chip as described in the above embodiment.

[0118] The present application also proposes a vehicle, comprising: a vehicle body, and the radar device as described above installed on the vehicle body.

[0119] The present application also provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the millimeter-wave radar signal processing method applied to a hardware accelerator integrated on a radar chip in the above-mentioned embodiments, and / or the millimeter-wave radar signal processing method applied to a digital signal processor integrated on a radar chip.

[0120] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0121] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other elements in the process, method, article or system including the element.

[0122] The above are only preferred embodiments of the present application, and the scope of the patent of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made by using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the scope of patent protection of the present application. The above are only preferred embodiments of the present application, and the scope of the patent of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made by using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the scope of patent protection of the present application.

Claims

1. A millimeter wave radar signal processing method, characterized in that: The method is applied to a hardware accelerator integrated on a radar chip, and the method comprises: Obtaining one-dimensional Fourier data corresponding to each detection target, wherein the one-dimensional Fourier data is obtained by converting the original received signal; When the digital signal processor extracts the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the last detected target, a two-dimensional Fourier transform is performed on the one-dimensional Fourier data corresponding to the current detected target; Obtaining two-dimensional Fourier data corresponding to the current detection target, and sending the two-dimensional Fourier data corresponding to the current detection target to the digital signal processor; Return to the step of performing a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target when the digital signal processor extracts the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the previous detection target, until the digital signal processor obtains the amplitude and phase values ​​corresponding to all detection targets.

2. The method according to claim 1, characterized in that The step of obtaining one-dimensional Fourier data corresponding to each detection target includes: Performing fast Fourier transform on the original received signal along the range dimension and the Doppler dimension respectively to obtain a range-Doppler spectrum diagram; Based on the range-Doppler spectrum diagram, a constant false alarm detection algorithm is used to perform target detection to obtain constant false alarm detection information; A plurality of detection targets are obtained according to the constant false alarm detection information, and one-dimensional Fourier data corresponding to each of the detection targets is determined.

3. The method according to claim 2, characterized in that The constant false alarm detection information includes: a distance index of each of the detection targets; The step of performing a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target includes: Perform a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the current detection target according to the distance index of the current detection target.

4. The method according to claim 2, characterized in that After the step of performing one-dimensional Fourier transform on the original received signal to obtain one-dimensional Fourier data corresponding to each of the detection targets, the method further includes: The range-Doppler spectrum diagram and the constant false alarm detection information are stored in a memory unit, so that the digital signal processor obtains the range-Doppler spectrum diagram and the constant false alarm detection information through the memory unit.

5. A millimeter wave radar signal processing method, characterized in that: The method is applied to a digital signal processor integrated on a radar chip, and the method comprises: When the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the next detection target, receiving the two-dimensional Fourier data corresponding to the current detection target; Extracting amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase value corresponding to the current detection target; Return to the step of receiving the two-dimensional Fourier data corresponding to the current detection target when the hardware accelerator performs a two-dimensional Fourier transform on the one-dimensional Fourier data corresponding to the next detection target, until the amplitude and phase values ​​corresponding to all the detection targets are obtained.

6. The method according to claim 5, characterized in that Before the step of extracting the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target, the method includes: Acquire a range-Doppler spectrum diagram and constant false alarm detection information, wherein the constant false alarm detection information includes: a Doppler index of each detected target; Correspondingly, the step of extracting the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target includes: The range-Doppler spectrum diagram corresponding to the current detection target is demodulated according to the code type information, and the amplitude and phase values ​​of the two-dimensional Fourier data corresponding to the current detection target are extracted according to the Doppler index of the current detection target to obtain the amplitude and phase values ​​corresponding to the current detection target.

7. The method according to claim 6, characterized in that The constant false alarm detection information also includes: a distance index of each of the detection targets, and after the step of obtaining the amplitude and phase values ​​corresponding to all the detection targets, the method includes: According to the distance index and Doppler index of each detection target, each detection target is screened by using a preset strategy to obtain a screened detection target; A preset angle measurement algorithm is used to perform angle measurement calculation on the amplitude and phase values ​​corresponding to each of the screened detection targets to obtain angle measurement results corresponding to each of the screened detection targets.

8. A storage medium, characterized in that: The storage medium stores a millimeter-wave radar signal processing program, which, when running, implements the steps of the millimeter-wave radar signal processing method according to any one of claims 1 to 4 and / or any one of claims 5 to 7.

9. A radar device, characterized in that: The radar device includes: a radar body, a hardware accelerator and a digital signal processor, the hardware accelerator is used to execute the millimeter-wave radar signal processing method as described in any one of claims 1-4, and the digital signal processor is used to execute the millimeter-wave radar signal processing method as described in any one of claims 5-7.

10. A vehicle, characterized in that: The vehicle comprises: a vehicle body, and the radar device according to claim 9 mounted on the vehicle body.

Images