Radar processing chain for frequency-modulated continuous wave radar systems

Abstract

To provide systems and methods for a radar processing chain for frequency-modulated continuous wave radar systems.SOLUTION: A transmitter 102 transmits a plurality of chirps, each comprising an electromagnetic radiation signal, at a region of interest. A receiver front-end 104 receives reflected electromagnetic radiation for each chirp and generates a time series of beat-signal samples for each chirp at each antenna of a plurality of antennas. A signal processor 106 detects objects within the region of interest by providing a frequency domain representation of each time series of beat-signal samples as sample values for a set of range bins representing respective distances from the receiver, correcting the sample values for each of the set of range bins to provide a set of clutter corrected samples for each range bin, and determining an angular spectrum for each of a subset of the set of range bins from the clutter corrected samples.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] This application relates generally to electrical systems, and more particularly to radar processing chains for frequency modulated continuous wave (FMCW) radar systems. [Background technology]

[0002] One form of distance measurement radar is based on frequency modulation, where the frequency of the returning signal is compared to the frequency of the transmitted signal. This technique can be used in continuous wave radars, for example in vehicular applications such as adaptive cruise control (ACC), emergency braking, pedestrian detection, and 360-degree sensing, as well as industrial applications such as building automation, people counting, and surveillance. In these systems, a carrier signal is frequency modulated in a predetermined manner to provide a chirp that is transmitted to the area of ​​interest. The received reflection from the area of ​​interest can be continuously compared to the transmitted chirp using a beat frequency modulator that generates a frequency tone representative of the time delay between the transmission of the chirp and the reception of the reflection. This can be determined from the known frequency modulation of the chirp. The distance traveled by the reflected signal can be determined from this time delay, which is twice the distance of the object from which the signal was reflected from the radar system. Summary of the Invention

[0003] According to one aspect of the disclosure, a radar processing chain for a frequency modulated continuous wave radar system is provided. A transmitter transmits a plurality of chirps, each comprising an electromagnetic radiation signal, in a region of interest and generates a time series of beat signal samples for each chirp at each antenna of a plurality of antennas. A receiver front end receives reflected electromagnetic radiation for each chirp. A signal processor detects any objects within the region of interest by providing a frequency domain representation of each time series of beat signal samples as sample values ​​for a set of range bins representing a respective distance from the receiver, correcting the sample values ​​for each of the set of range bins to provide a set of clutter-corrected samples for each range bin, and determining an angular spectrum for each of a subset of the set of range bins from the clutter-corrected samples.

[0004] According to another aspect of the disclosure, a method is provided. A plurality of chirps, each comprising a frequency modulated continuous wave electromagnetic radiation signal, are transmitted in a region of interest. Electromagnetic radiation reflected from the region of interest after each chirp is received at a receiver front end. The received electromagnetic radiation for each chirp is mixed with the transmitted chirp to provide a beat signal. The beat signal is sampled to provide a time series of beat signal samples for each chirp of the plurality of chirps at each antenna of the plurality of antennas. A frequency domain representation of each time series of beat signal samples is generated to provide a sample value for each of a set of range bins, each of which represents an associated distance from the receiver front end, for each chirp at each antenna. The frequency domain representation of the set of range bins is corrected to provide a set of clutter correction samples including a sample for each range bin and each chirp at each antenna. An angular spectrum is determined for a subset of the set of range bins from the clutter correction samples. Any objects within the region of interest are detected from the determined angular spectrum across the set of range bins.

[0005] According to yet another aspect of the disclosure, a method is provided for processing radar returns to detect objects within a region of interest. A time series of beat signal samples is generated for each chirp of a plurality of chirps at each antenna of a plurality of antennas. A frequency domain representation of each time series of beat signal samples is generated for each chirp at each antenna to provide a sample value for each of a set of range bins, each of which represents an associated distance from the radar sensor. The frequency domain representation of the set of range bins is corrected to provide a set of clutter-corrected samples including a sample for each range bin and each chirp at each antenna. An angular spectrum of a subset of the set of range bins is determined from the clutter-corrected samples. A location of at least one object within the region of interest is determined from the determined angular spectrum across the set of range bins. For a given range bin and chirp, spatial beamforming is applied to a -dimensional spatial signal block including all clutter-corrected samples for a given range bin of the set of range bins and a given chirp of the plurality of chirps across the plurality of antennas to provide a beamformed vector for the range bin. The given range bin is selected from the set of at least one range bin associated with at least one object. A Fast Fourier Transform (FFT) is applied to the beamformed signal blocks for the range bins to provide a Doppler spectrum vector from which Doppler information is extracted or at least one object associated with a given range bin is extracted. [Brief description of the drawings]

[0006] [Figure 1] Illustrates a frequency modulated continuous wave (FMCW) radar system.

[0007] [Diagram 2] 1 illustrates an example of signal processing components for processing a frequency modulated continuous wave (FMCW) radar signal.

[0008] [Diagram 3] 1 illustrates an example of a method for monitoring a region of interest.

[0009] [Figure 4] 1 illustrates an example of a method for processing received radar signals from a region of interest.

[0010] [Diagram 5] FIG. 5 is a schematic block diagram illustrating an example system of hardware components capable of implementing the example systems and methods disclosed in FIGS. 1-4, such as the digital signal processor illustrated in FIG. 2. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Radar processing chains have severe constraints in cluttered environments, such as determining the number and location of people in a region of interest. The systems and methods described herein provide a radar signal processing chain that includes clutter rejection algorithms and enables high-resolution direction of arrival (DoA) estimation and low-complexity Doppler estimation for frequency modulated continuous wave radar systems.

[0012] FIG. 1 illustrates a frequency modulated continuous wave (FMCW) radar system 100. The system 100 includes a transmitter 102 that transmits an FMCW electromagnetic radiation (EM) signal at a region of interest. It will be appreciated that the transmitter may include one or more antennas. In one implementation, the frequency modulated continuous wave EM signal is a “chirp” signal having a frequency that increases linearly with time, but it will be appreciated that the system 100 may utilize any suitable signal whose frequency varies with time in a known manner. The transmitter 102 may provide a set of chirp signals, called a frame. In one example, each frame includes 64 chirps, and the transmitter 102 transmits 20 frames per second. One or more receiver front ends 104 receive electromagnetic radiation reflected from the region of interest for each chirp at each of a plurality of antennas, mix the received electromagnetic radiation with the transmitted electromagnetic signal, sample the resulting beat signal to provide a time series of beat signal samples for each of the plurality of antennas, and provide the time series of beat signal samples to a signal processor 106. In one implementation, the multiple antennas are virtual antennas, with outputs derived via digital signal processing from the outputs of the physical transmit and receive antenna arrays. In such a case, a time series of beat signal samples is generated for each of the virtual antennas. An example of a radar system using a time division multiplexed multiple input multiple output (TDM-MIMO) implementation of a virtual antenna array is described in TI Application Report SWRA554A (see http: / / www.ti.com / lit / an / swra554a / swra554a.pdf), which is incorporated herein by reference.

[0013] One or more receiver front ends provide a time sequence of beat signal samples for each of the multiple antennas. Specifically, the transmitted FMCW signal is compared to the return signal to determine the frequency difference between the transmitted and returned signals. For example, the signals may be heterodyne mixed to generate a beat frequency from the two signals. Since the frequency modulation pattern is known, the change in frequency between the signals at a given time may provide the time of flight for that signal, and therefore the range information at which the returned signal was reflected.

[0014] The signal processor 106 evaluates the time series of beat signal samples to detect objects within the region of interest. It should be understood that the signal processor may be implemented, for example, as a dedicated microprocessor. The signal processor 106 provides a frequency domain representation of each time series of beat signal samples as a discretized representation of return strength over range, and as a value for a set of range bins representing distance from the radar sensor module. Thus, for each chirp signal, a set of range bin values ​​may be determined for each of the multiple antennas via a Fast Fourier Transform (FFT). In one implementation, the set of range bin values ​​is determined by averaging a set of signal samples from the chirp's beat signal to provide a DC component and subtracting the DC component from each sample to provide a set of DC compensated samples. The DC compensated samples are then subjected to a Fast Fourier Transform to provide a set of range bin values. In some applications, instead of using all samples received after a given chirp, a windowing operation may precede the DC compensation.

[0015] The signal processor 106 then corrects the frequency domain representation of the set of range bin values ​​to provide a clutter-corrected signal. Clutter correction may be applied, for example, by determining a direct current (DC) component as the average (e.g., mean) range bin value over the chirp in the frame for a given range bin and antenna, and subtracting the determined DC component from each range bin value associated with a given range bin and received at a given antenna. Thus, after a frame of chirps has been evaluated, the resulting data may be conceptualized as a three-dimensional data structure, where each sample value in the clutter-corrected signal represents a given chirp, range bin, and virtual antenna.

[0016] The signal processor 106 also uses the sample values ​​in the clutter-corrected signal to determine an angular spectrum for each range bin. In practice, the angular spectrum may be determined via application of a covariance-based direction-of-arrival determination algorithm, such as a Minimum Variance Distortion-Free Response (MVDR) algorithm, a Multiple Signal Classification (MUSIC) algorithm, or an Estimation of Signal Parameters by Rotational Invariance Techniques (ESPRIT) algorithm. The signal processor 106 may then examine the generated angular spectrum over a set of range bins to determine whether an object is present. In one implementation, the determined angular spectrum is represented as a range-azimuth spectrum matrix that represents the strength of the returned signal as a two-dimensional mapping of range and azimuth, referred to as a "range-azimuth profile" for the radar sensor module. This matrix may be searched for peak values, for example, via a Constant False Alarm Rate (CFAR) algorithm. The location of any detected object may then be stored in a non-transient memory 108 for later use, for example, by an automated system associated with the frequency-modulated continuous wave radar system 100.

[0017] FIG. 2 illustrates an example of a signal processing component 200 for processing a frequency modulated radar signal. For example, the signal processing component 200 may be used to provide the functionality of the signal processor 106 of FIG. 1. In the illustrated example, the signal processing component 200 receives samples representing either physical or virtual antennas from one or more receiver front ends. In the illustrated example 200, an array of virtual antennas is assumed to be derived from a set of physical transmit and receive antennas, but it will be understood that the multiple antennas may all be physical receive antennas representing beat signals between the transmitted chirp and the received signal from each chirp generated in a mixer and sampled in an analog-to-digital converter (ADC) during the chirp time. The samples from each receiver front end are processed in a respective range processing block 206.

[0018] Each range processing block 206 performs a Fast Fourier Transform (FFT) for an associated antenna to provide a set of values ​​representing the strength of the reflected signal from each of a number of distinct ranges from the antenna. In one implementation, a window function may be applied to the input samples of the FFT before generating the FFT. A direct current (DC) component in the input signal samples may be estimated by averaging the samples in each chirp and subtracted from each sample in the chirp. If a window function is used, any estimation and compensation for DC is applied after the window operation. In the illustrated implementation, the FFT of each chirp for each antenna is stored in a buffer as a set of range bin values. In each frame, multiple N c Chirp, number of virtual antennas N a , and a defined number of range bins N r If there is, then this is N r ×N c ×N a A three-dimensional data structure having entries is provided.

[0019] In cluttered environments, especially indoor environments, it can be difficult to detect objects with small radar cross section (RCS), such as humans and other life forms. Often, clutter sources with large RCS, such as building structures and furniture, can dominate the received signal. As a result, objects with small RCS can be buried in strong interference from these clutter sources and become difficult to detect. When the radar sensor is stationary, the clutter sources and all stationary objects do not exhibit a Doppler component, which can be exploited by the clutter rejection component 208 to improve the signal condition for the rest of the radar signal processing chain.

[0020] In one implementation, each clutter rejection component 208 may estimate the DC component of each range bin across the chirp in the frame. The DC component estimate is obtained by averaging the range FFT output samples across the chirp as follows: JPEG2025013859000002.jpg674Here, the range FFT output samples are {x n,k,p}, where x n,k,p represents the signal for the nth range bin in the kth chirp of the pth antenna, where n is a function of 1 to the number of range bins, N r where k is an integer between 1 and N, and k is the number of chirps in the frame. c where p is an integer between 1 and N, and n is the number of antennas in the system. a is an integer number.

[0021] Each range FFT output sample x n,k,p From the estimated DC component If you subtract TIFF2025013859000003.tif33, JPEG2025013859000004.jpg524. This is repeated for each range bin and for each antenna. Range processing and clutter removal operations are applied to each antenna and the resulting samples are stored as a set of clutter-corrected samples. The clutter-corrected samples are indexed using range index n, chirp index k, and antenna index p (of size N r ×N c ×N a The matrix can be conceptualized as a three-dimensional data structure.

[0022] The clutter-corrected samples from each antenna are provided to a Direction of Arrival (DoA) component 210, which uses the clutter-corrected samples to determine a DoA angular spectrum for each range bin. Collectively, the angular spectra form a range-azimuth spectrum. Specifically, the DoA component 210 performs high-resolution direction of arrival spectrum estimation based on the spatial covariance estimated within the radar frame. In the illustrated example, a minimum variance distortion-free response (MVDR) based Direction of Arrival (DoA) estimation approach is described. However, it will be appreciated that other covariance-based high-resolution DoA methods (e.g., MUSIC or ESPRIT) may be used as well as using the same radar processing signal chain.

[0023] The DoA component 210 first calculates the DoA ratio for each range bin as follows: Estimate spatial covariance for JPEG2025013859000005.jpg34. JPEG2025013859000006.jpg623, where x n,k is the N a is a dimensional space vector, which is formed by stacking samples across the virtual antennas, x n,k =[x n,k,1 ,x n,k,2 ,x n,k,Na ] T x Hdenotes the Hermitian transpose of the vector x. To reduce the amount of computation, we use the number of available chirps in the frame, N c A number of chirps, K, less than or equal to is used.

[0024] In the MVDR approach, also known as Capon beamforming in the illustrated implementation, a steering vector a(θ) for an azimuth angle θ is determined for a linear one-dimensional virtual receiver antenna array as follows: JPEG2025013859000007.jpg556 where d is the spacing between the antennas normalized to the wavelength.

[0025] The steering vector contains expected phase values ​​across the virtual antenna array for each hypothesized DoA angle, and represents the phase difference for each of the virtual receive antennas when the event radar radio frequency (RF) signal comes from an azimuth angle θ. The angular spectrum of the nth range bin is given by: JPEG2025013859000008.jpg521

[0026] The corresponding beamforming vector is given by: JPEG2025013859000009.jpg775

[0027] The DoA spectrum for each range bin is stacked in a matrix form, and the nth row is the nth range bin P n The DoA spectrum of (θ) is entered. This matrix, referred to herein as the range-azimuth spectrum matrix, represents the intensity of electromagnetic radiation reflected from a region of interest at a set of ranges and angles relative to the radar sensor.

[0028] In some situations, the spatial covariance estimate for a given range bin is singular or nearly singular, which can cause numerical stability issues when performing matrix operations (e.g., matrix inversion) as part of the DoA spectrum estimation. One way to avoid numerical stability issues is to add a diagonal loading factor α along the diagonal of the covariance estimate matrix as follows: n The trick is to add a small value called JPEG2025013859000010.jpg681

[0029] Diagonal loading factor α n is the noise variance estimate, JPEG2025013859000011.jpg34, and Using the constant β , JPEG2025013859000012.jpg35, where the product of the diagonal loading factors and the identity matrix is ​​referred to herein as the diagonal loading matrix. Specifically, in an environment that varies slowly over time compared to the time scale of the chirps in a radar frame, it can be assumed that the signal content of the clutter-corrected output does not change significantly between two adjacent chirps, but the noise content changes independently. This can be exploited to estimate the noise power or variance embedded in the radar signal for a given range bin n and virtual receive antenna p as follows: JPEG2025013859000013.jpg675, where x n,k,p are the clutter-corrected samples provided to the DoA component 210.

[0030] The noise variance estimate for range bin n may be determined by calculating the average noise variance over the virtual antennas as follows: JPEG2025013859000014.jpg575

[0031] The computational cost for DoA spectrum estimation can be reduced by performing DoA estimation only on range bins that are likely to have one or more objects. A one-dimensional constant false alarm rate (CRAR) algorithm can be used to determine good candidates for DoA spectrum estimation. In one implementation, a range profile vector s = [s1, s2, ..., s Nr ]'s nth element s n The CFAR algorithm is prepared by averaging the clutter correction sample values ​​over all chirps and all virtual receive antennas in the frame such that σ i is determined as follows: JPEG2025013859000015.jpg675

[0032] Once the range profile vector is determined, a one-dimensional CFAR algorithm may be performed on the range profile vector s to determine range bins that may contain the object, referred to herein as "detected range bins." In the CFAR algorithm, values ​​in the range profile vector that are peaks, i.e., rise above surrounding elements in the array (e.g., vector or matrix) and background levels, may be selected as detected range bins. The DoA spectrum estimation described above may be performed only on the detected range bins.

[0033] The range-azimuth spectrum matrix determined by the DoA component 210 is provided to an object detection component 212 that identifies the location of any objects within the region of interest. In one implementation, the object detection component 212 applies a two-dimensional CFAR algorithm to the range-azimuth spectrum matrix. Information related to the detected objects is provided, for example, by the range index n i , the angle index m in the matrix i , the power level, and the noise power level estimate are stored as i, where i is an index representing the i-th detected object.

[0034] Once a set of objects have been located, their positions in the range-azimuth-spectrum matrix are provided to the Doppler processing element 214, which applies spatial beamforming to the multiple streams of data from the virtual antennas and performs a Doppler FFT on the resulting beamformed data streams. In one embodiment, for the i-th detected object, the range bin n i For each chirp, each space vector x ni,k Then, spatial beamforming is applied to the signal block, resulting in a single stream of beamformed signal blocks, y i =[y i,k : k = 1, , N c ] is provided. JPEG2025013859000016.jpg677 where, JPEG2025013859000017.jpg830, where the superscript H denotes the Hermitian transpose of the vector.

[0035] From Equation 5, the beamforming weight vector w ni (θ mi It will be appreciated that the spatial covariance estimates for the particular range bins depend on the spatial covariance estimates. In one example, the resulting signal block y t A windowing operation can be applied to the resulting beamformed signal block y i The FFT operation is applied to obtain the Doppler spectrum vector Y i The Doppler spectrum vector Y i From the i-th detected object's Doppler information can be extracted. Then, all the information of the detected object, which may include range information, Doppler information, DoA angle information, and power and noise levels, is collected and stored in the buffer 216 as a form of "point cloud radar data".

[0036] In view of the structural and functional features described above, the example method may be better understood with reference to Figures 3 and 4. For simplicity of explanation, the example method of Figures 3 and 4 is shown and described as being performed sequentially, however, it should be understood and appreciated that the examples are not limited by the order shown, as some operations may occur multiple times and / or simultaneously in other examples, in different orders than shown and described herein. Moreover, it is not necessary that all of the actions described be performed to implement a method.

[0037] FIG. 3 illustrates an example of a method 300 for monitoring a region of interest. At 302, multiple chirps, each comprising a frequency modulated continuous wave (FMCW) electromagnetic radiation signal, are transmitted at the region of interest. At 304, after each chirp, electromagnetic radiation reflected from the region of interest is received at a receiver front end. It will be appreciated that the chirps transmitted at 302 may be transmitted from one or more antennas, and the reflected electromagnetic radiation may be received at multiple physical antennas. At 306, the received electromagnetic radiation for each chirp is mixed with the transmitted chirp to provide a beat signal, and at 308, the beat signal is sampled to provide a time series of beat signal samples for each chirp of the multiple chirps at each antenna of the multiple antennas. In one implementation, values ​​from the time series of beat signal samples for each chirp at each antenna are averaged to provide an estimated DC component of those values, and each time series of beat signal samples may be corrected for a DC by subtracting the estimated DC for that chirp and antenna pair from each of the time series of values ​​for a given chirp at a given antenna.

[0038] At 310, a frequency domain representation of each time series of beat signal samples is generated to provide sample values ​​for each of a set of range bins for each chirp at each antenna. Each range bin represents an associated distance from the receiver. In one example, the frequency domain representation is generated via a Fast Fourier Transform (FFT) on each of the values ​​in the time series. At 312, the frequency domain representation of the set of range bins is corrected to provide a set of clutter-corrected samples, including a sample for each range bin and each chirp at each antenna. In one implementation, the set of clutter-corrected samples may be generated by averaging samples over a subset of multiple chirps at each antenna, and an estimated direct current (DC) component of the value may be provided for each range bin at each antenna, such that a DC component is generated for each possible pair of antenna and range bin. The estimated DC component for each range bin and antenna pair is then subtracted from each of the values ​​in a given range bin for a given antenna.

[0039] At 314, an angular spectrum for a subset of the set of range bins is determined from the clutter-corrected samples. For example, the angular spectrum for each range bin may be determined by applying a minimum variance distortion-free response (MVDR) algorithm to the clutter-corrected samples. In one implementation, the subset of the set of range bins includes all of the range bins such that the subset is coextensive with the set of range bins. In another implementation, the subset of range bins is selected by generating a range profile vector that represents the total signal strength for each of the set of range bins across multiple chirps and multiple antennas, and determining from the range profile vector a set of range bins that are likely to contain the object as a subset of the set of range bins. For example, a one-dimensional constant false alarm rate algorithm may be applied to the range profile vector.

[0040] At 316, the location of any objects within the region of interest is determined from the angular spectrum determined over the set of range bins. In one implementation, the angular spectrum for a subset of the set of range bins can be represented as a range-azimuth spectrum matrix that represents the intensity of electromagnetic radiation reflected from the region of interest at a set of distances and angles relative to the radar sensor module. The location of any objects within the region of interest can be determined by applying a two-dimensional constant false alarm rate algorithm to the range-azimuth spectrum matrix. If there is location and other information available about the detected objects, it can be stored in a buffer at 318 for use by one or more other systems.

[0041] 4 illustrates an example of a method 400 for processing received radar signals in a region of interest. At 402, a time series of beat signal samples is generated for each chirp of a plurality of chirps at each antenna of a plurality of antennas. In one implementation, the time series of beat signal samples is generated by transmitting a plurality of chirps, each chirp including a frequency modulated electromagnetic radiation signal, at the region of interest and receiving at a receiver front end electromagnetic radiation reflected from the region of interest after each chirp. The received electromagnetic radiation from each chirp is mixed with the transmitted chirp to provide a beat signal, and the beat signal is sampled to provide a time series of beat signal samples for each chirp of the plurality of chirps at each antenna of the plurality of antennas.

[0042] At 404, a frequency domain representation of each time series of beat signal samples is generated to provide sample values ​​for each of a set of range bins for each chirp at each antenna. Each range bin represents an associated distance from the receiver. In one example, the frequency domain representation is generated via a Fast Fourier Transform (FFT) for each of the time series of values. At 406, the frequency domain representation of the set of range bins is corrected to provide a set of clutter correction samples, including a sample for each range bin and each chirp at each antenna. In one implementation, the set of clutter correction samples may be generated by providing an estimated direct current (DC) component of the values ​​for each range bin at each antenna by averaging samples for the set of range bins across a subset of multiple chirps at each antenna, such that a DC component is generated for each possible pair of range bins and chirps. The estimated DC component for each range bin and antenna pair is subtracted from each of the values ​​in a given range bin for a given antenna.

[0043] At 408, an angular spectrum for a subset of the set of range bins is determined from the clutter-corrected samples. For example, the angular spectrum for each range bin may be determined by applying a covariance-based direction of arrival determination algorithm to the clutter-corrected samples. In one implementation, the subset of the set of range bins includes all of the range bins such that the subset is coextensive with the set of range bins. In another implementation, the subset of range bins is selected by generating a range profile vector that represents the total signal strength for each of the set of range bins across multiple chirps and multiple antennas, and determining from the range profile vector a set of range bins that are likely to contain the object as a subset of the set of range bins. For example, a one-dimensional constant false alarm rate algorithm may be applied to the range profile vector.

[0044] At 410, any object in the region of interest is located from the determined angular spectrum across the set of range bins. At 412, spatial beamforming is applied to a Na-length spatial vector including all clutter-corrected samples for a given range bin of the set of range bins across the multiple antennas and a given chirp of the multiple chirps to provide a beamformed signal block for the range bin for the given range bin and chirp. It will be appreciated that the range bin may be selected from the set of range bins associated with any detected object. At 414, an FFT is applied to the beamformed signal block for the range bin to provide a Doppler spectrum vector. At 416, Doppler information for the object associated with the range bin is extracted from the Doppler spectrum vector. The position of any detected object, the extracted Doppler information, and other available information about the detected object may be stored in a buffer at 418 for use by one or more other systems.

[0045] FIG. 5 is a schematic block diagram illustrating an example system 500 of hardware components that may implement the example systems and methods disclosed in FIGS. 1-4, such as the digital signal processor 200 illustrated in FIG. 2. The system 500 may include various systems and subsystems. The system 500 may be a personal computer, a laptop computer, a workstation, a computer system, an appliance, an application specific integrated circuit (ASIC), a server, a server blade center, a server farm, and the like. In one implementation, the system 500 is a microprocessor, and it will be understood that in this implementation some of the components 502, 504, 506, 508, 510, 512, 514, 516, and 518 may be unnecessary. In one example, the digital signal processor 200 may be implemented as an integrated circuit. Examples of integrated circuit hardware that may be adapted for this purpose are described in TI data sheet SWRS203A, available at http: / / www.ti.com / general / docs / datasheetdiagram.tsp?genericPartNumber=AWRI624&diagramld=SWRS203A, and TI data sheet SWRS202A, available at http: / / www.ti.com / general / docs / datasheetdiagram.tsp?genericPartNumber=AWRI623&diagramld=SWRS202A, which are incorporated herein by reference.

[0046] The system 500 may include a system bus 502, a processing unit 504, a system memory 506, memory devices 508 and 510, a communication interface 512 (e.g., a network interface), a communication link 514, a display 516 (e.g., a video screen), and input devices 518 (e.g., a keyboard and / or a mouse). The system bus 502 may communicate with the processing unit 504 and the system memory 506. Additional memory devices 508 and 510, such as a hard disk drive, a server, a stand-alone database, or other non-volatile memory, may also communicate with the system bus 502. The system bus 502 interconnects the processing unit 504, the memory devices 506-510, the communication interface 512, the display 516, and the input devices 518. In some examples, the system bus 502 also interconnects additional ports (not shown), such as a Universal Serial Bus (USB) port.

[0047] The processing unit 504 may be a computing device and may include an application specific integrated circuit (ASIC). The processing unit 504 executes a sequence of instructions to implement the example operations disclosed herein. The processing unit may include a processing core.

[0048] Additional memory devices 506, 508, and 510 may store data, programs, instructions, database queries in text or compiled form, and any other information required to operate the computer. Memory 506, 508, and 510 may be implemented as a computer-readable medium (integrated or removable), such as, for example, a memory card, a disk drive, a compact disk, or a server accessible over a network. In some examples, memory 506, 508, and 510 may include text, images, video, and / or audio, some of which may be available in a human understandable form. Additionally or alternatively, system 500 may access external data sources or query sources via a communication interface 512 in communication with system bus 502 and communication link 514.

[0049] In operation, the system 500 may be used to implement one or more components of a diagnostic and decision support system according to the present disclosure. According to a particular embodiment, computer executable logic for implementing a composite application test system resides in the system memory 506 and one or more of the memory devices 508, 510. The processing unit 504 executes one or more computer executable instructions emanating from the system memory 506 and the memory devices 508 and 510. As used herein, the term "computer readable medium" refers to any medium that participates in providing instructions to the processing unit 504 for execution, and it will be understood that the computer readable medium may include multiple computer readable media each operably connected to the processing unit.

[0050] Modifications may be made to the exemplary embodiments described, and other embodiments are possible, within the scope of the invention.

Claims

1. A system comprising: a receiver coupled to an antenna, Receives radar signals based on reflections of the set of transmitted chirps; determining a set of beat samples of the radar signal based on the set of transmitted chirps; The receiver configured as above; a processor coupled to the receiver, performing a range Fast Fourier Transform (FFT) on the set of beat samples to determine a set of range bin values ​​for each chirp in the set of transmitted chirps; for each group of range bin values ​​corresponding to the set of range bin values, the group extending across the set of transmitted chirps, determining an average value for the group and adjusting each range bin value in the group based on the average value for the group to generate a set of adjusted range bin values; performing object detection on the radar signal based on the adjusted set of range bin values. The processor, Including, the system.

2. The system according to claim 1, The receiver, mixing the radar signal with a transmitted signal corresponding to the set of transmitted chirps to determine an analog beat signal; sampling the analog beat signal to generate the set of beat samples; The system is further configured to determine the set of beat samples by:

3. The system according to claim 1, The system, wherein the processor is further configured to determine Doppler information for objects detected by the object detection performed on the radar signal.

4. The system according to claim 1, The system, wherein the processor is further configured to adjust each range bin value of the group by subtracting it from an average value of the group.

5. The system according to claim 1, The system wherein the set of transmitted chirps is part of a frequency modulated continuous wave electromagnetic radiation signal.

6. The system according to claim 5, The system further includes a transmitter configured to transmit the frequency modulated continuous wave electromagnetic radiation signal.

7. A system comprising: A receiver circuit element coupled to a plurality of antennas, comprising: Receives radar signals based on reflections of the set of transmitted chirps; determining a series of beat samples of the radar signal for each antenna of the plurality of antennas based on the set of transmitted chirps; the receiving circuitry configured to a processor coupled to the receiving circuitry, determining an array of range bin values ​​for each chirp in the set of transmitted chirps; and performing a range Fast Fourier Transform (FFT) on each series of beat samples to determine a set of arrays of range bin values ​​for each antenna of the plurality of antennas; generating a set of arrays of adjusted range bin values ​​for each antenna of the plurality of antennas based on the series of beat samples; determining a direction of arrival spectrum for a subset of range bins associated with the set of arrays of adjusted range bin values; performing object detection on the radar signal based on the direction of arrival spectrum for the subset of range bins. The processor, Including, the system.

8. The system according to claim 7, the subset of range bins for the set of arrays of adjusted range bin values ​​is a first subset of range bins; the set of arrays of range bin values ​​surveyed is associated with a second subset of range bins that encompasses the first subset of range bins; The system, further configured to: determine the first subset of range bins by performing a constant false alarm rate (CFAR) algorithm on the first subset of range bins to determine the direction of arrival spectrum.

9. The system of claim 7, The system, wherein the processor is further configured to determine the direction-of-arrival spectrum by applying at least one of a Minimum Variance Distortionless Response (MVDR) algorithm, a Multiple Signal Classification (MUSIC) algorithm, and an Estimation of Signal Parameters Using Rotationally Invariant Techniques (ESPIRT) algorithm to the set of arrays of adjusted range bin values.

10. The system of claim 7, The system, wherein the multiple antennas are configured as a time division multiplexed multiple input multiple output (TDM-MIMO) array.

11. The system of claim 7, The system, wherein the receiving circuitry includes a plurality of receivers configured to be coupled to the plurality of antennas.

12. The system of claim 7, The system, wherein the processor is further configured to determine Doppler information by applying spatial beamforming to the set of arrays of adjusted range bin values.

13. A system comprising: transmit circuitry configured to transmit a first radar signal including a set of chirps; A set of receiving antennas, a receiver circuit element coupled to the set of receive antennas, the receiver circuit element comprising: receiving a second radar signal based on the first radar signal; determining a beat sample of the time series of the second radar signal for each antenna of the set of receive antennas; the receiving circuitry configured to a processor coupled to the receiving circuitry, for each time series of beat samples associated with a respective receive antenna of the set of receive antennas, performing a range Fast Fourier Transform (FFT) on the time series of beat samples to determine a set of range bin values ​​for each chirp of the set of chirps; for each group of corresponding range bin values ​​extending across the set of range bin values, determining an average value for the group; forming an index of surveyed range bin samples for the set of chirps and the set of receive antennas; adjusting each range bin value in the group based on the average value to generate a set of adjusted range bin values ​​for each antenna; performing object detection on the second radar signal based on the adjusted range bin samples. The processor, Including, the system.

14. A method comprising: receiving a signal based on a reflection of the transmitted chirp noise set; determining a set of beat samples associated with the set of transmitted chirps based on the signal; and performing a range Fast Fourier Transform (FFT) on the set of beat samples to determine a set of range bin values ​​for each chirp in the set of transmitted chirps; determining an average value for each group of corresponding range bin values ​​extending across the set of range bin values, adjusting each range bin value in the group based on the average value to generate a set of adjusted range bin values ​​for the group; determining a direction of arrival spectrum for a subset of range bins associated with the adjusted set of range bin values; performing object detection on the signal based on the determined direction of arrival spectrum; A method comprising:

15. The method of claim 14, comprising: determining the set of beat samples mixing said signal with a transmit signal corresponding to said set of transmitted chirps to determine an analog beat signal; sampling the analog beat signal to generate the set of beat samples; A method comprising:

16. The method of claim 14, comprising: The subset of range bins associated with the set of searched range bin values ​​includes range bins for a subset of the set of chirps.

17. The method of claim 14, comprising: The subset of range bins associated with the interrogated set of range bin values ​​is determined by applying a constant false alarm rate (CFAR) algorithm.

18. The method of claim 14, comprising: The method, wherein determining the direction of arrival spectrum includes applying at least one of a Minimum Variance Distortionless Response (MVDR) algorithm, a Multiple Signal Classification (MUSIC) algorithm, and an Estimation of Signal Parameters by Rotationally Invariant Techniques (ESPIRIT) algorithm to the set of arrays of adjusted range bin values.

19. The method of claim 14, comprising: The method further includes determining Doppler information for detected objects by performing object detection on the signals.

20. The method of claim 19, comprising: The method, wherein determining the Doppler information includes applying spatial beamforming to the set of adjusted range bin values.

Images