Coprime coded (CPC) doppler division multiplexing (DDM) MIMO radar method and system

Abstract

A co-prime coded DDM MIMO radar system, device, architecture and method is equipped with a reference signal generator (112) that generates a transmit reference signal; a plurality of DDM transmit modules (11) that generate, condition and transmit a plurality of transmit signals, each transmit signal having a co-prime coded (CPC) progressive phase offset that is different from the transmit reference signal; a receiver module (12) that receives target return signals reflected from the plurality of transmit signals by a target and generates a digital signal from the target return signals; and a radar control processing unit (20) configured to detect Doppler spectrum peaks in the digital signal, wherein the radar control processing unit includes a Doppler disambiguation module (25) configured with a CPC decoder to associate each detected Doppler spectrum peak with a corresponding DDM transmit module, thereby generating a plurality of transmitter-associated Doppler spectrum peak detections.

Description

Technical Field

[0001] This invention generally relates to radar systems and associated operating methods. In one aspect, the invention relates to an automotive radar system formed by a multiple-input multiple-output (MIMO) radar array. Background Technology

[0002] Radar systems can be used to detect the range, speed, and angle of nearby targets. With technological advancements, radar systems can now be applied to many different applications; for example, automotive radar safety systems increasingly use radar systems to detect changes in the surrounding environment, such as approaching another vehicle for blind spot detection or detecting leading vehicles to improve cruise control. Accurate radar detection is also an integral part of autonomous vehicle control systems. However, there are challenges in accurately detecting the position and movement of objects using radar systems that can be built into automobiles. A common method to improve the angular resolution of automotive radar systems is to implement a multiple-input multiple-output (MIMO) automotive radar system using multiple transmit and receive antennas. In a MIMO radar system, the virtual array consists of multiple array elements, which are equal to the product of multiple transmit and receive antennas. The increased aperture of MIMO radar compared to a single-transmitter system improves the ability to separate objects based on their direction of arrival (DOA). However, MIMO radar transmitters must emit orthogonal waveforms from multiple transmitters to separate the combined response at the receiver end.

[0003] Existing solutions for providing orthogonality include encoding the transmitted radar waveform, for example by modulating a continuous wave (FMCW) waveform at different times or with different center frequencies, or by changing the phase or amplitude of the FMCW waveform. In conventional automotive radar systems, particularly Doppler Division Multiplexing (DDM) MIMO automotive radar systems, the maximum Doppler frequency is often insufficient to prevent speed ambiguity caused by the overlap of Doppler spectra from multiple transmitting antennas. The ambiguity introduced by DDM MIMO systems is that each detected Doppler spectral peak must be correctly associated with its corresponding illumination transmitter, but if the target's radial velocity exceeds the budgeted maximum speed allocated to a transmitter, the spectral peak will appear in the portion of the spectrum allocated to another transmitter, leading to incorrect association. This problem is exacerbated for systems serving highly dynamic driving scenarios with many transmitters and limited Doppler spectral bandwidth.

[0004] Existing radar systems attempt to address these challenges by performing two measurements on each FMCW waveform, which has an increasing frequency "upward linear frequency modulation pulse" and a decreasing frequency "downward linear frequency modulation pulse." However, such solutions are overly complex because the upward and downward linear frequency modulation pulses need to be correlated with each other for radar processing. Other solutions use time-division (TD) multiplexing techniques to temporally separate LFM waveforms from different transmitters, thereby separating signals from different transmitters at each receive channel and constructing a virtual MIMO array. However, the number of transmitters that can be used with a TD-MIMO system is limited because the coherent dwell time (i.e., the duration for which the target's echo signal can be coherently integrated on a moving target) is typically finite. Another drawback of conventional TD-MIMO systems is that the longer duration of frames or linear frequency modulation pulse sequences can cause the maximum Doppler shift (or, in effect, the target's radial velocity) that can be measured without ambiguity to drop multiple times, further limiting the number of transmitters that can be used with a TD-MIMO system. Due to these limitations, existing MIMO automotive radar systems are typically limited to using a small number of transmitters (e.g., three) to construct a relatively small MIMO virtual array. In summary, given the performance, design, complexity, and cost constraints of existing radar system applications, current radar system solutions are extremely difficult to implement in practice, as achieving the performance advantages of larger radar sizes to achieve higher resolution radar imaging remains a challenge. Summary of the Invention

[0005] It should now be understood that a radar architecture, circuit, method, and system have been provided, wherein a reference signal generator is configured to generate a sequence of transmitted reference signal waveforms (e.g., linear frequency modulated pulse signals). Additionally, the radar system includes multiple Doppler Division Multiplexing (DDM) transmit modules connected and configured to generate, modulate, and transmit multiple transmitted signals on N transmit antennas, each transmitted signal having a progressive phase offset with a coprime code (CPC) different from the transmitted reference signal. In selected embodiments, each DDM transmit module includes a slow time phase offset shifter connected to generate a coprime-coded progressive phase offset signal from the transmitted reference signal. In selected embodiments, the slow time phase offset shifter may be implemented to provide a phase offset signal between 0 and 2... LCPC-encoded integer values ​​are used to introduce a CPC-encoded progressive phase offset into the L-position phase rotator of the transmit reference signal. Additionally, each DDM transmit module may include a signal encoder that uses signal conditioning and power amplification to encode the coprime-encoded progressive phase offset signal to generate one of a plurality of transmit signals, and transmits one of the plurality of transmit signals as an RF-encoded transmit signal via a transmit antenna. In other embodiments, the plurality of DDM transmit modules encode the transmit reference signal to define the zero radial velocity position of the DDM transmitter module, such that it has an interval value consistent with a coprime number-based value. The radar system also includes at least a first receiver module connected and configured to receive target return signals reflected from the plurality of transmit signals by at least one target and to generate a digital signal based on the target return signals. The radar system further includes a radar control processing unit configured to detect Doppler spectral peaks in a digital signal. This radar control processing unit includes a Doppler disambiguation module configured to have a CPC decoder to associate each detected Doppler spectral peak with a corresponding DDM transmitter module, thereby generating multiple transmitter-associated Doppler spectral peak detections. In a selected embodiment, the Doppler disambiguation module is configured to construct a binary FIR transmitter filter for each DDM transmitter module, having a tap delay following a unique coprime interval sequence associated with that transmitter module. In such embodiments, the Doppler disambiguation module may be configured to perform threshold detection on the Doppler spectrum generated from the digital signal and generate a binary detection sequence of Doppler spectral samples having a '1' only at the entry corresponding to the detected unit. Additionally, the Doppler disambiguation module may be configured to filter the binary detection sequence through each binary FIR transmitter filter to associate the spectral peak with the corresponding DDM transmitter module if the corresponding filter output exceeds a predefined threshold, thereby identifying the transmitter-associated detection. Additionally, the Doppler disambiguation module can be configured to, for each transmitter-associated detection, evaluate the variance of the magnitude of the Doppler spectrum sample matching the tap delay position in the binary FIR transmitter filter against an evaluation threshold to confirm that each transmitter-associated detection has a variance equal to or lower than the evaluation threshold. In selected embodiments, the radar control processing unit can also be configured to construct a MIMO virtual array based on multiple transmitter-associated Doppler spectrum peak detections.

[0006] In another form, a radar system architecture and its operation method are provided. In the disclosed method, a linear frequency modulated (LFM) pulse signal is generated. Additionally, multiple DDM transmit signals are generated from the LFM pulse signal by multiple DDM transmitter modules, each DDM transmit signal having a coprime coded (CPC) phase offset different from the LFM pulse signal. In a selected embodiment, the DDM transmit signals are generated by encoding the LFM pulse signal to define the zero radial velocity positions of the multiple DDM transmitter modules, such that they have interval values ​​consistent with coprime numbers. In other embodiments, the DDM transmit signals are generated by applying multiple phase shifts to the LFM pulse signal using corresponding multiple multi-bit slow-time phase offset shifters to generate multiple CPC asymptotic phase offset signals from the LFM pulse signal, each CPC asymptotic phase offset signal having a different CPC phase offset from the LFM pulse signal. In such embodiments, each of the multi-bit slow-time phase offset shifters may be implemented with the ability to provide phase shifts between 0 and 2. LCPC-encoded integer values ​​are used to introduce a progressive phase shift of CPC-encoded phase into an L-bit phase rotator of the LFM linear frequency modulated pulse signal. In the disclosed method, one or more receiving antennas of the receiver module receive target return signals reflected from multiple DDM transmitted signals via a target. The disclosed method further includes mixing the target return signal with the LFM linear frequency modulated pulse signal at the receiver module to generate an intermediate frequency (IF) signal, and then converting the IF signal into a digital signal using an analog-to-digital converter at the receiver module. In the disclosed method, the digital signal is then processed to detect Doppler spectral peaks in the digital signal. Finally, the disclosed method applies a CPC decoder to associate each detected Doppler spectral peak with a corresponding DDM transmitter module, thereby generating multiple transmitter-associated Doppler spectral peak detections. In a selected embodiment, the CPC decoder can be applied by constructing a binary FIR transmitter filter for each DDM transmitter module, having a tap delay that follows a unique coprime interval sequence associated with the DDM transmitter module. In such embodiments, a CPC decoder can be applied by performing threshold detection on the Doppler spectrum generated from a digital signal, and then performing CPC decoding to generate a binary sequence of Doppler spectrum samples with '1' only at the entry corresponding to the detected cell. Alternatively, applying the CPC decoder may include filtering the binary detection sequence through each binary FIR transmitter filter to associate spectral peaks with the corresponding DDM transmitter module if the corresponding filter output exceeds a predefined threshold, thereby identifying the transmitter-associated detection. Additionally, the disclosed method may, for each transmitter-associated detection, evaluate the variance of the magnitude of the Doppler spectrum samples matching the tap delay positions in the binary FIR transmitter filter against an evaluation threshold to confirm that each transmitter-associated detection has a variance equal to or lower than the evaluation threshold. The disclosed method may also construct a MIMO virtual array based on multiple transmitter-associated Doppler spectrum peak detections.

[0007] In yet another form, radar architectures, circuits, methods, and systems are provided. In the disclosed system, multiple transmitter modules are configured to simultaneously transmit multiple Doppler divided-multiplexed (DDM) linear frequency modulated pulse waveforms in a single transmission frame having a fixed pulse repetition frequency (PRF) rate. Each DDM linear frequency modulated pulse waveform is encoded using a programmable multi-bit phase rotator with different coprime coding (CPC) phase offsets, such that the Doppler spectral signals from different transmitter modules can be separated in a single transmission frame when there is overlap between the spectra from different transmitter modules. Additionally, at least a first receiver module is connected and configured to receive target return signals reflected from the multiple DDM linear frequency modulated pulse waveforms by at least one target and generate a digital signal based on the target return signals. Furthermore, a radar control processing unit is configured to detect Doppler spectral peaks in the digital signal and apply CPC decoding to associate each detected Doppler spectral peak with a corresponding transmitter module, thereby generating multiple transmitter-associated Doppler spectral peak detections. In a selected embodiment, the radar control processing unit is configured to construct a binary FIR transmitter filter for each transmitter module, having a tap delay that follows a unique coprime interval sequence associated with the transmitter module. Additionally, the radar control processing unit may be configured to perform threshold detection on the Doppler spectrum generated from the digital signal and generate a binary detection sequence of Doppler spectrum samples having a '1' only at the entry corresponding to the detected unit. The radar control processing unit may also be configured to filter the binary detection sequence through each binary FIR transmitter filter to associate spectral peaks with the corresponding DDM transmitter module if the corresponding filter output exceeds a predefined threshold, thereby identifying the transmitter-associated detection. Finally, the radar control processing unit may be configured to, for each transmitter-associated detection, evaluate the variance of the magnitude of the Doppler spectrum samples matching the tap delay positions in the binary FIR transmitter filter against an evaluation threshold to confirm that each transmitter-associated detection has a variance equal to or lower than the evaluation threshold. Attached Figure Description

[0008] The invention and its many objectives, features and advantages will be understood when the following detailed description of preferred embodiments is considered in conjunction with the following figures.

[0009] Figure 1 This is a simplified schematic block diagram of a linear frequency modulated pulse DDM MIMO radar system according to a selected embodiment of the present disclosure.

[0010] Figure 2 This is a timing diagram showing the linear frequency modulated pulse transmission schedule of a uniform DDM MIMO radar system.

[0011] Figure 3 It is a curve of the Doppler spectrum of a uniform DDM MIMO radar system, in which the maximum radial velocity of the target does not exceed the allocated spectrum budget, thus producing a clear correlation between the target peak and the transmitting antenna.

[0012] Figure 4 This is a graph of the Doppler spectrum of a uniform DDM MIMO radar system, where the target's maximum radial velocity exceeds the allocated spectral budget, resulting in a clear correlation between the target peak and the transmitting antenna.

[0013] Figure 5 This is a timing diagram illustrating a linear frequency modulated pulse transmission schedule of a DDM MIMO radar system using a non-uniform Doppler segmentation scheme employing coprime coding, according to a selected embodiment of the present disclosure.

[0014] Figure 6 This is a graph of the CPC Doppler spectrum of a non-uniform DDM MIMO radar system, where CPC decoding technology can be used to assign the detection peak of each receiver to each transmitter even when there is overlap between the spectra of different transmitters.

[0015] Figure 7 A simplified flowchart is shown, illustrating the logic of forming a virtual large MIMO radar array using CPC DDM technology.

[0016] Figure 8 This is a simplified graphical depiction of the physical radar aperture and MIMO virtual array aperture provided by 3-TX and 4-RX antennas according to a selected embodiment of the present disclosure.

[0017] Figure 9 A spectral graph of a 3-TX CPC DDM MIMO radar system according to a selected embodiment of the present disclosure is depicted, wherein three targets are simulated to illustrate the CPC disambiguation process.

[0018] Figure 10 A spectral graph of a 3-TX CPC DDM MIMO radar system is plotted, simulating three targets and three spurious Doppler signals to illustrate the CPC disambiguation process that produces CPC decoding errors.

[0019] Figure 11 A spectral graph of a 3-TX CPC DDM MIMO radar system according to a selected embodiment of the present disclosure is depicted, wherein three targets and three spurious Doppler signals are simulated to illustrate the CPC disambiguation process, which produces CPC decoding errors corrected by a variance checking process. Detailed Implementation

[0020] A coprime-coded (CPC) Doppler split-multiplexing (DDM) MIMO radar system, hardware circuitry, system, architecture, and method are described to eliminate overlapping Doppler spectra from multiple transmitters by encoding Doppler split-MIMO waveforms using coprime-based zero radial velocity frequency spacing, allowing for stable and explicit reconstruction of Doppler spectrum detection from individual transmitters. In a selected embodiment, CPC coding is implemented in individual LFM DDM MIMO transmitters using a programmable phase rotator in each transmit channel to perform progressive phase coding over a slow time (i.e., the time of a linear frequency modulated pulse), such that the spacing value (between two adjacent zero radial velocity Doppler bands in the number of bins) is consistent with a coprime value. By applying a progressive phase shift to the linear frequency modulated pulse signal, the amount of progression between transmitters uses a coprime-based value, and each transmitted signal has a coprime-coded phase offset different from the transmitted reference signal. As a result, the slow-time derivative of the slow-time phase change (i.e., the slope of the asymptotic phase change function or linear phase curve over the linear frequency modulation pulse time) is the amount of Doppler frequency shift generated by phase coding. By controlling the slope (the amount of advancement), the placement of zero radial velocity on the Doppler spectrum for the transmitter can be controlled. In the case of transmitting different CPC-coded LFM signals on each transmit channel, the receiver can use a CPC decoder to associate Doppler peaks with their corresponding transmitters by matching the peak occurrences in the Doppler spectrum with the unique peak occurrence position codes of each transmitter in the decoding vector, thereby disambiguating the DDM MIMO waveform. In selected embodiments, the received signal processing steps may include: FIR filtering, which is used to directly separate the Doppler spectral peaks of each transmitter by defining the CPC decoding vector; detecting Doppler peaks; and performing CPC decoding by performing threshold detection on the Doppler spectrum of the received radar signal and generating a binary sequence of Doppler spectral sample lengths with '1' only at the entry corresponding to the detected unit. Furthermore, by evaluating the variance of the filter-matched spectral samples against a predefined threshold, a variance checking process can be applied to further reduce erroneous associations, ensuring that associations are marked as true when the variance is equal to or below the predefined threshold. This method allows additional transmitters to share the finite Doppler spectrum obtained from the maximum achievable pulse repetition frequency (PRF) rate in a fast linear frequency modulated pulse radar system, enabling the construction of large MIMO virtual arrays and higher angular resolution performance. Consequently, even when the spectra of individual transmitters highly overlap, the Doppler spectral signals from different transmitters can be effectively separated within a single transmission frame.By providing hardware and software solutions for disambiguating the overlapping Doppler spectra of multiple transmitters using coprime encoding and decoding, the disclosed LFM DDM MIMO radar system and method effectively provide a MIMO virtual array with an aperture many times larger than the combined total physical aperture, thereby achieving better sensitivity, finer angular resolution and lower false detection rate.

[0021] In the context of this disclosure, it should be understood that radar systems can be used as sensors in a variety of different applications, including but not limited to automotive radar sensors for road safety systems such as Advanced Driver Assistance Systems (ADAS) and Autonomous Driving (AD) systems. In such applications, radar systems are used to measure the radial distance, relative radial velocity, and angular information of a reflecting target, and are characterized by performance criteria such as angular resolution (the minimum distance between two given-size targets that the radar can distinguish and separate from each other at a resolution unit of the same distance and rate of change of distance (or radial velocity), sensitivity, false detection rate, etc. Typically, frequency modulated continuous wave (FMCW) modulated radar is used to identify the distance, velocity, and / or angle of radar targets such as cars or pedestrians by transmitting linear frequency modulated (LFM) waveforms from multiple transmitting antennas, such that reflected signals from the radar target are received and processed at multiple receiving antennas to determine the radial distance, relative radial velocity, and angle (or orientation) of the radar target. However, in the case of current automotive designs, vehicles may include multiple radar transmitters capable of operating independently of each other. While different coding domains (time, frequency, and code) exist for distinguishing individual transmitters, selected embodiments of this disclosure relate to LFM waveform transceivers configured to perform Doppler Partition Multiplexing (DDM) MIMO operation by having each transmitter transmit identical frequency ramped LFM pulse signals having specified phase offset values ​​on the individual LFM pulses encoded into the entire LFM pulse sequence to signals separated on the spectrum of different transmitters in the Doppler domain, such that the receiving channel can clearly detect each signal and thus construct a virtual MIMO array.

[0022] To understand the design and operation of a DDM MIMO radar system in context, refer now to Figure 1This diagram depicts a simplified schematic block diagram of a linear frequency modulated pulse (LFM) DDM-MIMO automotive radar system 100, which includes an LFM DDM-MIMO radar device 10 connected to a radar microcontroller unit (MCU) 20. In selected embodiments, the LFM DDM-MIMO radar device 10 may be embodied as a line replaceable unit (LRU) or modular component designed for rapid replacement in operating locations. Similarly, the radar MCU 20 may be embodied as a line replaceable unit (LRU) or modular component. Although a single or monostatic LFM DDM-MIMO radar device 10 is shown, it should be understood that additional distributed radar devices can be used to form a distributed or multistatic radar. Furthermore, the depicted radar system 100 may be implemented as an integrated circuit, wherein the LFM DDM-MIMO radar device 10 and the radar MCU 20 are formed by separate integrated circuits (chips) or by a single chip, depending on the application.

[0023] Each radar device 10 includes one or more transmit antenna elements TX that are respectively connected to one or more radio frequency (RF) transmitter (TX) units 11 and receiver (RX) units 12. i and receiving antenna element RX j For example, each radar device (e.g., 10) is shown as including individual antenna elements (e.g., TX) connected to three transmitter modules (e.g., 11) and four receiver modules (e.g., 12), respectively. 1，i RX 1，j However, these numbers are not limiting, and other numbers are also possible, such as four transmitter modules 11 and six receiver modules 12, or a single transmitter module 11 and / or a single receiver module 12. Each radar device 10 also includes a linear frequency modulated pulse generator 112, which is configured and connected to supply linear frequency modulated pulse input signals to the transmitter modules 11. For this purpose, the linear frequency modulated pulse generator 112 is connected to receive input program and control signals 110, including a reference local oscillator (LO) signal, a linear frequency modulated pulse start trigger signal, and a program control signal. A linear frequency modulated pulse signal 113 is generated and transmitted to multiple transmitters 11, wherein phase encoding is applied using a phase rotator 114, which is controlled by the program control signal 110 generated by the MCU 20. While phase shifter 114 provides a uniform phase shift between transmitters 11, programmable phase rotator 114 also provides each transmitter 11 with the ability to implement progressive phase shifts using non-uniform coding techniques, such as CPC coding techniques described more fully below. At each transmitter 11, the phase-coded linear frequency modulated pulse signal is then filtered at RF conditioning module 115 and fed to the corresponding transmit antenna TX.1，i And amplification is performed at power amplifier 116 before radiation. This is achieved by using each transmitting antenna to TX. 1，i The transmitter element 11 is used to transmit a progressively phase-shifted sequence of linear frequency modulated pulse signals. Each transmitter element 11 operates in a Doppler split multiplexing manner with other transmitter elements because these transmitter elements are programmed to transmit the same waveform simultaneously on a phase-separated timetable.

[0024] Through the transmitter antenna unit TX 1，i TX 2，i The transmitted radar signal can be reflected by the target, and a portion of the reflected radar signal reaches the receiver antenna unit RX at radar device 10. 1，i At each receiver module 12, the received (RF) antenna signal is amplified by a low-noise amplifier (LNA) 120 and then fed to a mixer 121, where the received (RF) antenna signal is mixed with a transmitted linear frequency modulated pulse signal generated by an RF conditioning unit 115. The resulting intermediate frequency signal is fed to a first high-pass filter (HPF) 122. The resulting filtered signal is fed to a first variable gain amplifier 123, which amplifies the signal before feeding it to a first low-pass filter (LPF) 124. This re-filtered signal is fed to an analog-to-digital converter (ADC) 125 and output as a digital signal D1 by each receiver module 12. In this way, the receiver module 12 compresses target echoes with various delays into multiple sinusoidal frequencies, the frequencies of which correspond to the round-trip delay of the echo.

[0025] In radar system 100, radar MCU 20 can be connected and configured to supply input control signals 110 to radar device 10 and receive digital output signals D1 generated by receiver module 12 therefrom. In selected embodiments, radar MCU 20 includes radar controller processing unit 21, which may be embodied as a microcontroller unit (MCU) or other processing unit configured and arranged for signal processing tasks, such as, but not limited to, calculation of target identification, target range, target velocity, and target direction, and generation of control signals.

[0026] To control the transmitting module 11, the radar controller processing unit 21 may be configured, for example, to generate transmitter input signals 110, such as program, control trigger, reference LO signal, calibration signal, spectrum shaping signal (e.g., ramp generation in the case of FMCW radar), and to receive data signals, sensor signals, and / or register programming or state machine signals for RF (radio frequency) circuit enable sequences. Specifically, the radar controller processor 21 may be configured to use a progressive phase shift and transmit LFM linear frequency modulated pulses for the transmitting antenna TX. 1，i RX 1，j Communication between them is used to program the transmitter module 11 to operate in DDM mode using the transmitter input signal 110. In a selected embodiment, the radar controller processor 21 is configured to use CPC-coded progressively phase-shifted LFM linear frequency modulated pulses 113 by programming a programmable phase rotator 114 before transmission. In the case where each transmitter module 11 transmits a transmit channel signal having a different CPC-coded LFM signal generated by the programmable slow time phase shifter 114, the receiver module 12 can adjust the target return signal to generate a digital domain signal D1, which is processed by the radar MCU 20 to separate and identify the CPC-coded transmit channel signal.

[0027] At the receiver module, a digital output signal D1 is generated from the target return signal for digital processing by radar modules 22-26 to construct and accumulate a multiple-input multiple-output (MIMO) array vector output, thereby forming a MIMO aperture for calculating the graph generated by angle estimation (e.g., spatial or angular FFT) module 27 and the target trajectory generated by target tracking module 28. Specifically, the digital output signal D1 may be processed by one or more Fast Fourier Transform (FFT) modules, such as Fast Time (Range) FFT module 22 which generates a Range Linear Frequency Modulated Pulse Antenna Cube (RCAC) and Slow Time (Doppler) FFT module 23 which generates a Range-Doppler Antenna Cube or Map (RDM). The RDM output is then passed through one or more detection modules 24, such as a Constant False Alarm Rate (CFAR) detection module, to obtain Range-Doppler Peak Detection (DET'NS). As described more fully below, the range-Doppler peak detection can be processed by the Doppler disambiguation module 25 to associate the Doppler peaks with the corresponding transmitters using a CPC decoder to match the peak occurrences in the Doppler spectrum with the unique peak occurrence position codes of each transmitter, thereby generating transmitter-associated detections (TX-ASSOCIATEDDET'NS), thus disambiguating the DDM MIMO waveform. The transmitter-associated detection peaks can be further processed at the spatial angle estimation module 27 and the target tracking module 28, where the resulting target trajectory output 29 is sent to other automotive computing devices or user interface devices for further processing or display.

[0028] In a selected embodiment, the Doppler disambiguation module 25 is configured to construct or access a binary FIR filter for each transmitter, having a tap delay that follows a unique coprime interval sequence associated with the transmitter, and then perform threshold detection on the Doppler spectrum of the received radar signal to produce a binary sequence of Doppler spectrum sample lengths with '1' only at the entry corresponding to the detected cell. The resulting binary detection sequence is filtered with each transmitter filter from the transmitter filters, and a spectral peak is associated with the transmitter if the corresponding filter output exceeds a predefined threshold. To improve performance in dense target environments, each association is analyzed by evaluating the variance of the Doppler spectrum samples matching the tap delay position to confirm the association if the variance is not greater than the predefined threshold. Once the Doppler peak is correctly associated with the starting transmitter module 11,

[0029] To understand the operation of a conventional DDM MIMO radar system in context, refer now to Figure 2This is a timing diagram 200 illustrating the linear frequency modulated pulse transmission schedule of four transmitters 201-204 using a uniform Doppler segmentation scheme. As depicted, each transmitter TX1-TX4 is programmed to simultaneously transmit a sequence of DDM linear frequency modulated pulse waveforms 201-204 within a single radar transmission frame. Each transmitter (e.g., TX1) transmits linear frequency modulated pulse waveforms (e.g., 201A, 201B) at a fixed and uniform pulse repetition frequency (PRF) rate, said rate being the reciprocal of the pulse repetition interval (e.g., PRF = PRI). -1 ) or the reciprocal of the linear frequency modulated pulse interval (e.g., PRF = CIT) -1 Additionally, each transmitter encodes each linear frequency modulation (LFM) pulse with an additional progressive phase shift using a phase rotator in the front-end circuitry. As a result of this progressive phase shift encoding, each received LFM pulse originating from each of the different transmitters TX1-TX4 effectively has a different zero radial velocity Doppler shift, and individual target detections can be associated with the correct starting transmitter, which is necessary for the proper operation of subsequent MIMO virtual array construction.

[0030] Based on the following equation, the position of zero radial velocity is controlled by a progressive phase shift of the linear frequency modulated pulse applied to each transmitter:

[0031]

[0032] Where f zrv，i It is the zero radial velocity Doppler shift frequency of transmitter i, and where A i It is the progressive phase shift (in radians) between two adjacent linear frequency modulated pulses. For example, in a 4-TX DDM MIMO system, each transmitter can be assigned the following progressive phase shift, and thus obtain a zero radial velocity frequency:

[0033] TX1: A1 = 0 → f zrv，1 =0[Hz]

[0034]

[0035]

[0036]

[0037] exist Figure 2In the example, this result is depicted as the first transmitter TX1 encoding its linear frequency modulated pulse waveform 201 with a progressive phase shift of 0 degrees, the second transmitter TX2 encoding its linear frequency modulated pulse waveform 202 with a progressive phase shift of 90 degrees, the third transmitter TX3 encoding its linear frequency modulated pulse waveform 203 with a progressive phase shift of 180 degrees, and the fourth transmitter TX4 encoding its linear frequency modulated pulse waveform 204 with a progressive phase shift of 270 degrees. In such DDM MIMO radar systems using uniform Doppler segmentation, a target can be definitively associated with the correct transmitter by determining its position within the segmented Doppler spectrum, provided that the target has a true Doppler shift within ±PRF / (2N) before the application of DDM, where N is the number of DDM transmitters.

[0038] To illustrate the principle of Doppler peak detection and transmitter correlation, refer now to Figure 3 The graph 300 depicts the Doppler spectrum as a function of the arc frequency (and Doppler shift) of a uniform DDM MIMO radar system, where the target's maximum radial velocity does not exceed the allocated spectral budget, thus creating a clear correlation between the target peak and the transmitting antenna. As depicted, each transmitter TX1-TX4 has an allocated spectral portion 311-314 effectively centered at the corresponding zero radial velocity frequencies (e.g., 0, π / 2, and 3π / 2). Therefore, each target measurement 301-304 with a true Doppler shift within ±PRF / (2N) of the zero radial velocity frequencies of each transmitter TX1-TX4 can be clearly associated with the correct transmitter by determining its position within the partitioned Doppler spectrum 300. This result holds true as long as the target's radial velocity falls within the allocated spectral budget portion 311-314 for each transmitter TX1-TX4.

[0039] In the case of existing 76-81 GHz fast linear frequency modulated pulse (LFM) automotive radar front-end monolithic microwave integrated circuits (MMICs), the fastest LFM pulse signal is limited by a LFM pulse interval time (CIT) of not less than approximately 15 microseconds. Therefore, the maximum definitive Doppler radial detection speed is ±65 m / s (or ±234 km / hr), and the entire range of detectable speeds is then divided into N parts for N transmitters to perform definitive DDM operation. As can be clearly seen, in high-dynamic driving scenarios where the target and radar are traveling at speeds of 100 km / hr or faster, a Doppler spectral bandwidth of approximately 470 km / hr is only sufficient for a very small number of transmitters. While significantly shortening the CIT to approximately microseconds could support a larger number of transmitters without introducing ambiguity, such ultrashort LFM pulses introduce additional cost and complexity to the radar system. Instead of radar systems that support ultra-short linear frequency modulation pulses or larger spectral budget allocations, ambiguity will appear in the DDM Doppler spectrum when the radial velocity of the target exceeds its allocated spectral budget, causing spectral peaks from the first transmitter to appear in the portion allocated to the other transmitter, resulting in incorrect correlation.

[0040] To illustrate how the correlation between the transmitter and spectral peaks can break down, refer now to Figure 4The graph 400 depicts the Doppler spectrum as a function of the arc frequency or Doppler shift of a uniform DDM MIMO radar system, where the target's maximum radial velocity exceeds the allocated spectral budget, resulting in an ambiguous correlation between the target peak and the transmitting antenna. The share of the CPC Doppler spectrum 400 from each transmitter is indicated by solid lines 401 (for TX1), dashed lines 402 (for TX2), dashed lines 403 (for TX3), and dashed lines 604 (for TX4). As depicted, each transmitter TX1-TX4 has an allocated spectral portion 411-414 effectively centered at the corresponding zero radial velocity frequencies (e.g., 0, π / 2, and 3π / 2). However, the detections 401-404 of transmitters TX1-TX4 encroach on the allocated spectral portions of different transmitters, making them impossible to correlate explicitly. Specifically, the detections 401A and 401C for the first transmitter TX1 are present in the allocated spectrum 411A and 411B of the first transmitter TX1, but one or more additional detections for the first transmitter TX1 (e.g., 401B circled in gray) are present in the allocated spectrum of a different transmitter (e.g., spectrum 414 for the fourth transmitter TX4). Similarly, the detections 402B and 402C for the second transmitter TX2 are present in the allocated spectrum 412 of the second transmitter TX2, but one or more additional detections for the second transmitter TX2 (e.g., detection 402A circled in gray) are present in the allocated spectrum of a different transmitter (e.g., spectrum 411 for the first transmitter TX1). Similarly, the detections 403B and 403C for the third transmitter TX3 are present in the spectrum 413 allocated to the third transmitter TX3, but one or more additional detections for the third transmitter TX3 (e.g., detection 403A circled in gray) are present in the spectrum allocated to different transmitters (e.g., spectrum 412 for the second transmitter TX2). Finally, the detections 404B and 404C for the fourth transmitter TX4 are present in the spectrum 414 allocated to the fourth transmitter TX4, but one or more additional detections for the fourth transmitter TX4 (e.g., detection 404A circled in gray) are present in the spectrum allocated to different transmitters (e.g., spectrum 413 for the third transmitter TX3).

[0041] As transistor peak detection encroaches on the spectrum allocated to other transistors, errors arising when associating the detection with the transmitter make it impossible to construct a MIMO virtual array output without correcting the association. Therefore, a method for disambiguating the overlapping Doppler spectrum of a DDM transmitter is needed.

[0042] To address other limitations known from conventional solutions and those of those skilled in the art, reference is now made to... Figure 5This is a timing diagram 500 illustrating a linear frequency modulated pulse transmission schedule for a DDM MIMO radar system with three transmitters 501-503 using a non-uniform Doppler segmentation scheme employing coprime coding, according to a selected embodiment of this disclosure. As depicted, each transmitter TX1-TX3 is programmed to simultaneously transmit a series of DDM linear frequency modulated pulse waveforms 501-503 with a fixed and uniform pulse repetition frequency (PRF) rate, said rate being the reciprocal of the pulse repetition interval (e.g., PRF = PRI). -1 ) or the reciprocal of the linear frequency modulated pulse interval (e.g., PRF = CIT) -1 Prior to transmission, each linear frequency modulated pulse (LFM) pulse is encoded with an additional progressive phase shift using the CPC encoding technique disclosed herein, employing a programmable phase rotator in the front-end circuitry. By using progressive phase-shift CPC encoding, the Doppler spectra of transmitters TX1-TX3 are effectively allocated according to the coprime number theory, allowing the Doppler spectral signals of different transmitters to be separated into a single transmission frame even when the spectra of the individual transmitters severely overlap. By encoding the DDM MIMO waveform using a coprime zero radial velocity frequency spacing, the Doppler spectrum detection of each transmitter can be stably and definitively reconstructed to be associated with the correct transmitter, thereby enabling more transmitters to share a finite Doppler spectrum limited by the maximum realizable PRF. Therefore, each received LFM pulse originating from each different transmitter TX1-TX3 effectively has a different zero radial velocity Doppler frequency shift, and each target detection can be associated with the correct starting transmitter, which is necessary for the correct operation of subsequent MIMO virtual array construction.

[0043] exist Figure 5In the example, the distinguished CPC coding scheme is described as follows: the first transmitter TX1 encodes its linear frequency modulated pulse waveform 501 with a progressive phase shift of 0 degrees; the second transmitter TX2 encodes its linear frequency modulated pulse waveform 502 with a progressive phase shift of 17 / 64×PRF = 95.625 degrees; and the third transmitter TX3 encodes its linear frequency modulated pulse waveform 503 with a progressive phase shift of 43 / 64×PRF = 241.875 degrees. Therefore, for zero radial velocity detection from the first transmitter TX1, the first transmitter TX1 has an allocated spectral portion centered at 0×PRF = 0 degrees. Similarly, for zero radial velocity detection from the second transmitter TX2, the second transmitter TX2 has an allocated spectral portion centered at 17 / 64×PRF = 95.625 degrees. Finally, for the zero radial velocity detection from the third transmitter TX3, which has an allocated spectral portion centered at 43 / 64 × PRF = 241.875 degrees, the received Doppler-segmented MIMO waveforms can be processed using CPC decoding by encoding the individual DDM transmitters TX1-TX3 using coprime-based zero radial velocity frequency intervals to correctly associate the individual transmitter Doppler spectral detections with the correct transmitter. It will be noted that the arrangement of the zero radial velocities generating the interval values ​​{17, 26, 21} is a coprime number in PRF / 64 [Hz].

[0044] To better understand selected embodiments of this disclosure, reference is now made to Figure 6The graph depicts the CPC Doppler spectrum 600 of a non-uniform DDM MIMO radar system with detection peaks from three transmitters TX1-TX3. Even with overlap between the spectra used for different transmitters, CPC decoding technology can be used to correctly assign the receiver to or associate it with the corresponding transmitter. The share of the CPC Doppler spectrum 600 from each transmitter is indicated by a solid line 601 (TX1), a gray line 602 (TX2), and a dashed line 603 (TX3). In this case, each transmitter TX1-TX3 has an allocated spectrum 611-613 of the full PRF range in Hz (or 2π radians), with each spectrum centered at a frequency corresponding to the zero radial velocity (e.g., 0, 2πx(17 / 64), and 2πx(43 / 64)). With this arrangement of the CPC Doppler spectrum 600 and the disclosed decoding method, the receiver can associate peak detection 601A in the spectrum 611A / B of the first transmitter with the first transmitter TX1. Additionally, the receiver can associate peak detection 602A in the spectrum 612A / B of the second transmitter with the second transmitter TX2, and the receiver can associate peak detection 603A in the spectrum 613A / B of the third transmitter with the third transmitter TX3.

[0045] from Figure 6 As shown and referred to below Figure 9-11The described example demonstrates that using coprime or quasi-prime spacing between the zero-velocity points of the transmitters on the Doppler spectrum 600 prevents peak detections 601A-603A from being associated with incorrect transmitters, even when the spectra overlap, thereby achieving a clear association between each detected peak and its corresponding transmitter. At the receiver, this Doppler spectrum arrangement can be physically detected by observing the phase shift of each transmitted linear frequency modulated pulse applied to each transmitter, by using a spectrum analyzer to detect the shift in the center frequency of the transmitter's entire transmission frame, or by observing the Doppler spectrum generated by the output of each range cell over a slow time. In selected embodiments, the radar receiver can be configured to perform different signal processing steps that apply FIR filtering to the received radar signal waveform to separate the Doppler spectral peaks of each transmitter, thereby eliminating any complex if / then dynamic decision logic operations. Additionally, digital processing steps at the radar receiver can provide additional plausibility checks by evaluating the variance of the filter-matched spectrum samples to further reduce the false association rate. The disclosed FIR filtering method effectively separates the received Doppler spectrum signals from different transmitters and associates them with the corresponding transmitter in a single radar transmission frame, even when the spectra of the various transmitters severely overlap (reaching the full 2π range). This represents an improvement over conventional methods that require multiple frames and complex decision logic to complete any disambiguation task. Furthermore, the performance of the disclosed FIR filtering method scales proportionally with the number of transmitters, as the filter length increases linearly with the number of transmitters. In fact, compared to conventional methods that are non-scalable and whose performance decreases with the number of transmitters, longer FIR filters have better discriminative power to compensate for the effects of denser spectra.

[0046] To provide additional details that enable a better understanding of selected embodiments of this disclosure, reference is now made to... Figure 7 It depicts a simplified flowchart 700 illustrating the logic of forming a virtual large MIMO radar array using CPC DDM technology. In an example embodiment, Figure 7 The control logic and methods shown can be implemented as hardware and / or software on a host computing system, processor, or microcontroller unit, which includes a processor and a memory for storing programming control code for constructing and operating a large virtual MIMO radar array by introducing a CPC-based phase-off modulation signal into a reference linear frequency modulated pulse signal to achieve separation of the transmitter signal in the Doppler spectrum frequency domain for the output of each range unit.

[0047] The process begins (step 701), for example, when a radar system begins the process of sensing the position and movement of one or more target objects using one or more transmitted radar signals transmitted on multiple transmit antennas. To generate the transmitted radar signal, the radar system first generates a reference linear frequency modulated (LFM) pulse signal (step 702), for example, by periodically modulating the transmitted radar signal with frequency and / or phase shift. For example, for an automotive frequency modulated continuous wave (FMCW) radar, the reference LFM pulse signal may be generated as a linear frequency modulated (LFM) waveform distributed to multiple transmit channel circuits respectively associated with multiple transmit antennas.

[0048] At step 703, a reference linear frequency modulated (LFM) pulse signal is applied to multiple slow-time phase shifters to generate coprime-coded (CPC) phase-shifted LFM pulse signals for multiple transmit channels. In selected embodiments, the phase-shifting step can be implemented by applying the reference LFM pulse signal to multiple phase shifters, each controlled by a phase-shift control signal, to carefully impose the design allocation of the transmitter's Doppler spectrum according to the principles of coprime number theory. Alternatively, one of the transmit channel circuits can be connected to directly receive the reference LFM pulse signal without any phase-shift modulation. Instead of uniformly distributing the Doppler spectrum portion (as in a typical DDM implementation), the processing at step 703 can use phase shifters to introduce a non-uniform, progressive phase shift into the reference LFM pulse signal at the transmit channel to control the position of the transmitter's zero radial velocity, such that the interval value (between two adjacent zero radial velocity Doppler bands in the number of bands) is consistent with a coprime value in some units.

[0049] With 64 linear frequency modulation pulses (i.e., N) c =64, where N c In an example embodiment of a 3-TX DDMMIMO radar system (where the number of linear frequency modulated pulses is the number of pulses), the phase shift control signal supplied to the phase shifter can be selected to generate intervals [17, 26, 21] in units of PRF / 64 [Hz] (meaning modulus Nc in the wrapped spectrum), such that the zero radial velocity band index of the three transmitters (based on 0, from 0 to N) c -1) These are [0, 17, 43] in PRF / 64 [Hz]. In this example, the interval values ​​are coprime numbers in PRF / 64 [Hz] because their prime factorizations (e.g., 17 = 1*17, 26 = 1*2*13, and 21 = 1*3*7) have no common factors other than 1, so they can only be factored by prime numbers. Using coprime-based intervals ensures that each Doppler detection peak can be correctly associated with its transmitter in later steps.

[0050] In selected embodiments, progressive phase coding over a slow time (i.e., the linear frequency modulation pulse time) can be achieved by using a phase rotator for each transmit channel. For an L-bit phase rotator, it is possible to... radians or Programming up to 2 radians with quantization step size L One phase shift position. For example, the step size of a 6-bit phase rotator is... In radians or 5.625 degrees. In this case, the minimum phase progression between two linear frequency modulated pulses is 5.625 degrees, and the maximum phase progression between two linear frequency modulated pulses is... Or 354.375 degrees. Generally, the k-th phase level setting can be defined as... Where k = 0, 1, 2, ... 2 L -1, and therefore, the phase offset of the m-th linear frequency modulated pulse is Where m = 1, 2, ... N c And mod{·,2π} represents the modulus 2π operator. In this case, setting the k-th phase series will make the doppler spectrum... Zero radial velocity is observed.

[0051] To perform an example where zero radial velocity is located in the [0, 17, 43] band of the 64-band long Doppler spectrum, the setting for k is also [0, 17, 43] for a 6-bit phase rotator. The progressive phase coding of the first transmitter TX1 over a slow time will be evaluated for the second transmitter TX2 at k=0, for the third transmitter TX3 at k=43 (i.e., 0° between linear frequency modulation pulses for TX1; 95.625° between linear frequency modulation pulses for TX2; and 241.875° between linear frequency modulation pulses for TX3).

[0052] As will be understood, a 6-bit phase rotator provides possible CPC-coded integer values ​​between 0 and 63. If the number of Doppler bands is 64, the band spacing can be encoded using coprime numbers. If there are more than 2... L Each frequency band (e.g., 64 frequency bands in the 6-bit case) can be based on the equation Calculate the number N of the zero radial velocity frequency band. zrv，k , where N Dop It is the number of Doppler bands (possibly greater than N). c ), and in 2 L Without dividing kN Dop In the case of N zrv，k It may not be an integer. By choosing N Dop So that N zrv，kIt is an integer, and zero radial velocity can be aligned with the center of the Doppler band. When N Dop Greater than 2 L At that time, the interval between zero radial velocities may no longer be exactly coprime, but this approximation should not affect the effectiveness of the CPC coding scheme.

[0053] As described above, CPC coding can be used to achieve coprime spacing between zero-velocity points of the transmitter on the Doppler spectrum. However, it will be understood that other non-uniform spacing techniques, including but not limited to quasi-coprime spacing techniques, can be used to prevent peak detection from being incorrectly associated with the transmitter. Alternatively, some pseudo-random spacing at the zero radial velocity position can be used instead of the coprime method, and good performance can still be achieved. Using coprime spacing ensures optimal decoding stability, but acceptable good performance can still be achieved by using non-coprime but pseudo-random spacing.

[0054] At step 704, the CPC linear frequency modulated pulse signal is conditioned and amplified using Doppler split-multiplexing technology for transmission on the corresponding transmit channel circuit. In a selected embodiment, this is performed by transmit channel circuits, each including an RF conditioning module (which may include filtering, leveling, applying additional phase coding, or applying RF switching to the output of the corresponding phase shifter) and a power amplifier (which amplifies the RF conditioning module output for transmission on the corresponding transmit antenna).

[0055] At step 705, reflected CPC-coded linear frequency modulated (LFM) pulse signals from different transmit channels are received, conditioned, and converted for digital processing at the receiver. In a selected embodiment, one or more receiving antennas at the receiver module receive the target return as a (RF) antenna signal from the transmitted CPC-coded phase-offset reference LFM pulse signal waveform for subsequent amplification, for example, by using a low-noise amplifier to generate an amplified RF signal from the target return. Additionally, the received transmit channel signal reflected by the target (i.e., the target signal) is mixed with the reference LFM pulse signal at the receiver to generate an intermediate frequency (IF) signal. In a selected embodiment, the mixing step can be implemented by applying the reference LFM pulse signal to a receiver module mixer, which is also connected to receive the received target signal for mixing with the reference LFM pulse signal, thereby generating the IF signal. Furthermore, the IF signal can be adjusted for digital conversion. In a selected embodiment, the conditioning process includes feeding the IF signal to a high-pass filter and amplifying the filtered signal with a variable gain amplifier before feeding it to a low-pass filter, thereby generating a re-filtered signal. Additionally, the refiltered, regulated IF signal can be fed to a high-speed analog-to-digital converter (ADC) with a digital signal output suitable for digital processing.

[0056] At step 706, digital processing is applied to separate the reflected transmit channel signals of each transmitter using a DDM MIMO CPC disambiguation signal processing step. While any suitable CPC disambiguation signal processing step can be used at step 706, in a selected embodiment, the processing may begin at step 711 by constructing a binary FIR transmitter filter for each transmitter with a tap delay that follows a unique coprime interval sequence associated with the transmitter module, thus defining the CPC decoding vector.

[0057] In operation, Doppler peaks can be associated with their transmitters by matching the occurrence of peaks in the Doppler spectrum with a unique peak occurrence location code for each transmitter. (Combined) Figure 6 In the example shown, the transmitter's code (in frequency band number) is:

[0058] For transmitter #1, C1 = [0, 17, 17 + 26] = [0, 17, 43].

[0059] For transmitter #2, C2 = [0, 26, 26 + 21] = [0, 26, 47], and

[0060] For transmitter #3, C3 = [0, 21, 21 + 17] = [0, 21, 38]

[0061] Because the values ​​{17, 26, 21} are coprime, and the values ​​{43, 47, 38} are also coprime, the three decoded vectors (or codes) C1, C2, and C3 guarantee an explicit match for the peak occurrences of each transmitter. Since the property of coprime numbers is that the sum of coprime numbers is also coprime, the code values ​​are also coprime.

[0062] Continuing this example, the three peak position codes can be represented in binary sequence format to form three binary FIR filter decoding vectors. In this case, the bit positions corresponding to the indicator positions of each code in each sequence are set to '1', while the rest are set to '0'. For example, the three codes C1, C2, and C3 can be written in binary form as follows:

[0063] C1=[10000000000000000100000000000000000000000001]

[0064] C1=[1←16“0”s→1←25“0”s→1]

[0065] C2=[1000000000000000000000000010000000000000000000001]

[0066] C2=[1←25“0”s→1←20“0”s→1]

[0067] C3=[100000000000000000000100000000000000001]

[0068] C3=[1←20“0”s→1←16“0”s→1]

[0069] The digital disambiguation processing step 706 may further include a Doppler peak detection step 712, which performs threshold detection on the Doppler spectrum of the received radar signal to generate a binary sequence of Doppler spectrum sample lengths with '1' only at entries corresponding to detected cells that meet the threshold requirement. While any suitable Doppler peak detection processing step may be used, the processing at step 712 can be implemented by creating a binary sequence based on the peak detection results for each detected range band and for the Doppler spectrum of each receiver. The peak amplitude (complex value) should be buffered for subsequent MIMO array construction and angle estimation processes, and optionally buffered for additional variance checks as described below.

[0070] At step 713, CPC decoding is performed to disambiguate the Doppler spectral signals from different transmitters by filtering the binary sequence (created at step 712) with a binary FIR filter (created at step 711) to identify correlations where the filter output exceeds a predefined threshold, thereby restoring the correct peak-transmitter correlation. In a selected embodiment, the CPC decoding process begins with any Doppler-detected peak to test N by matching the subsequent N-1 peak positions. c CPC decoding vector assumptions. If a peak detection exists at any of the peak positions indicated by all N-1 codes, a candidate match is declared. Due to the uniqueness of the codes, only one assumption will produce a perfect match.

[0071] While any suitable CPC decoding procedure can be used at step 713, a selected embodiment of the CPC decoding technique can be implemented by correlating the binary detection sequence with each of the binary FIR filters in the decoding loop and detecting a match at the correlator position when the correlator output is equal to N. In other words, for the filter output of the first transmitter (e.g., TX1), if the correlator output value N is in the peak position, then the peak is associated with the first transmitter, and so on. Generally, for cases where N is large, a decoder filter output threshold lower than N can be used to allow for missed detections. For cases where N is small, such as 3-TX, the threshold should be set to N (i.e., 3).

[0072] As an optional step 714, variance checking processing can be applied to the evaluation associations against one or more evaluation criteria (identified at step 713) to further reduce the false association rate. As indicated by the dashed lines, the variance checking processing step can be omitted or skipped in the disclosed sequence 700. However, selected embodiments of the variance checking processing step can be used to improve performance in dense target environments by evaluating the variance of Doppler spectrum samples matching the tap delay position against a predefined variance checking evaluation threshold for each identified association, and then marking each identified association as “true” if the variance is not greater than the predefined evaluation threshold.

[0073] In the variance checking process, the peak amplitude of the matching position identified by CPC decoding step 713 can be further analyzed to ensure that the match is not due to overlap caused by false target detection or crowded target scenes. While any suitable evaluation process can be used, in selected embodiments, a simple test statistic, such as the variance or standard deviation of the N peak amplitudes (in absolute value), can be calculated and checked against a user-defined variance checking evaluation threshold. If the test statistic is below the user-defined evaluation threshold, a final association decision is made. Variance checking process step 714 can be used to handle low-probability events where the detection of other targets at the encoded position results in a falsely detected match. In such cases, a large variance is likely to occur because the amplitude of other targets is unlikely to be the same as that of legitimate targets. Conversely, the amplitude should be consistent for a true match, thus ensuring a higher confidence level in the decoded output by ensuring the variance is below the evaluation threshold.

[0074] At step 707, a virtual MIMO array is constructed based on the transmitter-associated detections extracted from the reflected transmit channel signals originating from different transmit channels. In a selected embodiment, the output from the processing at step 714 (which associates the Doppler detection peak with the initial DDM transmitter) is used to construct a MIMO virtual array using monostatic and / or bistatic radar principles. For this purpose, the estimated Doppler frequency shift or radial velocity can be determined by the band position relative to the zero radial velocity band of each transmitter, and receiver outputs with the same radial velocity (i.e., the same number of relative bands) are grouped according to a standard MIMO virtual array construction process to form the MIMO array outputs.

[0075] At step 708, the MIMO virtual array output is processed by a range estimation process, a Doppler estimation process, and an angle estimation process, and a target map is generated to identify the range value, Doppler value, and angle value of each detected target. The range estimator, Doppler estimator, and angle estimator are typically based on Fast Fourier Transform (FFT) processors and Discrete Fourier Transform (DFT) processors. More advanced spectrum estimators may also be used for angle processing, including but not limited to Multi-Signal Classifier (MUSIC) and Signal Parameter Estimator (ESPRIT) processors via rotation invariance techniques. In selected embodiments, the radar controller processor may be configured to generate paired range(r) values, Doppler values, and angle values ​​identifying each detected / target object. Mapping data between values ​​and angle (θ) values.

[0076] To provide additional details that enable a better understanding of selected embodiments of this disclosure, reference is now made to... Figure 8 This is a simplified schematic description 800 of a sample radar device with 3-TX antennas 801 and 4-RX antennas 802, depicting the physical radar apertures 801, 802 and the MIMO virtual array aperture 803. In the depicted radar device, the distributed arrangement of the three transmit antennas TX1-TX3 801 and the four receiver antennas RX1-RX4 802 is positioned and distributed relative to each other, as shown in the figure. Without loss of generality, the physical antennas are shown as being positioned linearly, but they can be arranged non-linearly and / or interleaved. The depicted MIMO virtual array antenna 803 is formed by transmitting radar signals from the three transmit antennas T1-T3 received at the receiver antennas R1-R4. On the left side of the MIMO virtual array antenna 803, the MIMO virtual array element VR... 1，1 VR 1，2 VR 1，3 VR 1，4 Generated by receiving antennas R1-R4, which receive radar transmitted signals radiated by the first transmitting antenna T1. The MIMO virtual array element VR is located in the middle of the MIMO virtual array antenna 803. 2，1 VR 2，2 VR 2，3 VR 2，4 Generated by receiving antennas R1-R4, which receive radar transmitted signals radiated by the second transmitting antenna T2. To the right of the MIMO virtual array antenna 803, the MIMO virtual array element VR... 3，1 VR 3，2 VR 3，3 VR 3，4Generated by receiving antennas R1-R4, which receive radar transmitted signals radiated by a third transmitting antenna T3. Therefore, the MIMO virtual array antenna 803 has more elements than the physical arrays of the transmitting and receiving antennas 801, 802, and occupies a larger (wider) area. Since angular resolution is inversely proportional to aperture size, the MIMO virtual array aperture 803 provides improved angular resolution.

[0077] By transmitting a CPC-coded linear frequency modulated pulse signal from the 3-TX antenna 801, the target return received at the 4-RX antenna 802 can be processed using CPC decoding to associate the Doppler peak detection samples from the target return with the initiating transmitter, and then used to construct a 12-element virtual array antenna VR. i，j 803. Next, the 12-element array spatial measurement output vector is constructed and processed. 804 is used to estimate the angle of arrival (AoA) of one or more targets. This process is then repeated for the next range band of interest until all range bands of interest have been processed.

[0078] To provide additional details that enable a better understanding of selected embodiments of this disclosure, reference is now made to... Figure 9 The diagram 900 depicts the spectrum of a 3-TX CPC DDM MIMO radar system, simulating three targets to illustrate the CPC disambiguation process. In the depicted example, a total of 256 CPC-coded linear frequency modulated pulses are transmitted by three DDM transmitters, and the simulated range cell output sample containing these 256 linear frequency modulated pulses from three targets illuminated by the three DDM transmitters is processed to produce the observed co-coded DDM spectrum 920. For reference purposes, a real target-transmitter correlation generated with the same CPC coding is depicted in the top diagram 910. In both cases, the CPC encoder is programmed such that the zero radial velocity positions of the three transmitters TX1, TX2, and TX3 are located at k = [0, 68, 152] (or equivalently at interval values ​​of [68, 84, 104]), based on coprime interval values ​​[17, 21, 26] with a scale factor of 4 (or equivalently, based on the zero radial velocity positions [0, 17, 38] with a scale factor of 4). The observed coprime-coded DDM spectrum 920 also depicts detected peaks (e.g., 921-923) above the detection threshold 924, where the detected peaks are marked with commas. A comparison with the reference coprime-coded DDM spectrum 910 reveals numerous false detections (e.g., 925-927) in the observed coprime-coded DDM spectrum 920 due to noise.

[0079] As disclosed herein, the CPC decoding process can filter out erroneous detections by using a CPC decoding vector FIR filter. The decoding results are shown in a disambiguation filter output graph 930, which illustrates the correlator output for each transmitter with different lines for the first transmitter 931 (dotted line), the second transmitter 932 (dashed line), and the third transmitter 933 (solid line). Correct associations are identified for each transmitter by finding that the correlator output value is close to the correction value N (e.g., 3 in this case). In the disambiguation filter output graph 930, the solutions identified above the detection threshold line 934 are marked with the numbers 1, 2, and 3 to indicate the transmitters associated with the Doppler peaks. In this case, all Doppler detections are correctly associated with the starting transmitter.

[0080] To provide additional details that enable a better understanding of selected embodiments of this disclosure, reference is now made to... Figure 10 The diagram 1000 depicts the spectrum of a 3-TX CPC DDM MIMO radar system, simulating three targets and three spurious Doppler signals to illustrate the CPC disambiguation process that produces CPC decoding errors. In the depicted example, a total of 256 CPC-coded linear frequency modulated pulses are transmitted by three DDM transmitters, and the simulated range cell output samples of the 256 linear frequency modulated pulses are processed to produce a co-coded DDM spectrum 1010 with simulated slow-time range band outputs containing three targets. In addition to the target signals associated with the first transmitter (e.g., 1011), the second transmitter (e.g., 1012), and the third transmitter (e.g., 1013), the simulated spectrum 1010 also includes three spurious Doppler signals 1014, 1015, and 1016 (marked with '◆') added to obfuscate the decoding process. Therefore, the DDM spectrum plots 1020, 1030, and 1040, associated with coprime codes of transmitters Tx1, Tx2, and Tx3 respectively, show erroneous associations in the decoder output at detections 1021 (in plot 1020 for the first transmitter), 1031 (in plot 1030 for the second transmitter), and 1041 (in plot 1040 for the third transmitter). Without further processing, these erroneous associations 1021, 1031, and 1041 will be used to construct an incorrect MIMO array output.

[0081] To resolve issues caused by incorrect associations, refer to the following: Figure 11The diagram 1100 depicts a spectral graph of a 3-TX CPC DDM MIMO radar system according to a selected embodiment of the present disclosure, wherein three targets and three pseudo-Doppler signals are simulated to illustrate the CPC disambiguation process, which produces CPC decoding errors corrected by a variance checking process. Similarly, a coprime-coded DDM spectral graph 1110 is generated by simulating the transmission of 256 CPC-coded linear frequency modulated pulses from three DDM transmitters, thus producing range cell output samples with simulated time-range band outputs containing three targets. In addition to the target signals associated with the first transmitter (e.g., 1111), the second transmitter (e.g., 1112), and the third transmitter (e.g., 1113), the simulated reference CPC DDM spectral graph 1110 also includes three pseudo-Doppler signals 1114, 1115, and 1116 (marked with '◆') added to obfuscate the decoding process. To filter out erroneous associations, the vector variance readings of the transmitter filters for transmitters Tx1, Tx2, and Tx3, plotted in graphs 1120, 1130, and 1140, depict the variance of the magnitudes of samples corresponding to the positions where the decoded vector or correlator indicates '1'. As shown in each graph 1120, 1130, and 1140, a threshold is used to detect association peaks with sufficiently small variances. In this way, erroneous associations tending to have variances higher than acceptable are filtered out during the process. For example, the detection threshold 1127 in the vector variance reading 1120 of the first transmitter filter is used to identify peaks 1123-1126 (indicated by) the final associations with variances lower than the detection threshold 1127. (labeled), and used to filter out peaks 1121-1122 with variances higher than the detection threshold 1127 (using... (Marking). Additionally, the detection threshold 1137 in the vector variance reading 1130 of the second transmitter filter is used to identify the final associated peaks 1133-1136 (using...) with variances below the detection threshold 1137. (label), and filter out peaks 1131-1132 with variances higher than the detection threshold 1137 (using... (Marking). Finally, the detection threshold 1147 in the vector variance reading 1140 of the third transmitter filter is used to identify the final associated peaks 1143-1146 (using) variances below the detection threshold 1147. (labeled), and used to filter out peaks 1141-1142 with variances higher than the detection threshold 1147 (using... (Mark). In this way, the correct result can be restored.

[0082] As disclosed herein, selected embodiments of the disclosed CPC phase-offset DDM MIMO radar system offer several enhancements compared to conventional radar systems. Beyond the construction of a maximum MIMO array for transmitting linear frequency modulated (LFM) waveforms in automotive frequency modulated continuous wave (FMCW) radar, the disclosed radar system can utilize the RF front-end and signal processing blocks of existing radar designs without significant modifications, thereby minimizing the cost of developing new solutions. For example, the disclosed CPC-encoded LFM pulse programming and processing method can be implemented on current and next-generation radar front-end MMICs and radar microcontroller / processor units (MCUs) to achieve full DDM MIMO capability without relying on ultrashort LFM pulse radar systems. Furthermore, due to the nature of coprime numbers, the disclosed use of coprime values ​​to encode different transmitters with non-uniform spacing ensures a much higher probability of correctly associating Doppler peaks with the starting transmitter. Another advantage is the stability of the disclosed CPC encoding technique, which scales proportionally with the number of transmitters as the code length increases, whereas conventional methods cannot scale stably because individual interruptions may not be strongly distinguishable and ambiguity may still occur. For example, Jansen's U.S. Patent No. 2020 / 049812A1 teaches a coding scheme that simulates a DDM spectrum with uniform spacing using a binary phase shifter with a Walsh-Hadamard binary phase coding sequence. Four uniform zero-velocity transmitters are coded using the Walsh-Hadamard scheme, leaving one transmitter position empty and supporting three transmitters performing DDM MIMO radar processing. Using this conventional method, decoding capability is much weaker because using suboptimal Hadamard codes only produces uniform spacing, resulting in highly similar codes. Worse still, the Doppler spectrum will contain spurious responses (false peaks) due to the simulation of asymptotic phase shifts using binary codes. This disclosure, however, uses a multi-bit phase rotator to directly encode the signal to provide non-uniform spacing, thereby enabling differentiation between transmitter associations.

[0083] The foregoing description of specific embodiments has described benefits, other advantages, and solutions to problems. However, these benefits, advantages, solutions to problems, and any elements that may make any benefit, advantage, or solution occur or become more significant should not be construed as essential, necessary, or fundamental features or elements of any or all claims. As used herein, the terms "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not only include those elements, but may include other elements not expressly listed or not inherent to such process, method, article, or apparatus.

Claims

1. A radar system, characterized in that, include: A reference signal generator is configured to generate a transmit reference signal comprising a waveform sequence; Multiple Doppler split-multiplexing (DDM) transmitter modules are connected and configured to generate, modulate, and transmit multiple transmit signals on N transmit antennas, each transmit signal having a coprime coded CPC phase offset different from the transmit reference signal; At least a first receiver module is connected and configured to receive a target return signal reflected from the plurality of transmitted signals by at least one target and to generate a digital signal based on the target return signal; as well as A radar control processing unit is configured to detect Doppler spectral peaks in the digital signal. The radar control processing unit includes a Doppler disambiguation module configured to have a coprime-coded CPC decoder to associate each detected Doppler spectral peak with a corresponding Doppler split-multiplexing (DDM) transmitter module, thereby generating multiple Doppler spectral peak detections associated with the transmitter. Each Doppler split-multiplexing (DDM) transmitter module includes: A slow time phase shifter is connected to generate a coprime-coded progressive phase shift signal from the transmitted reference signal; as well as A signal encoder uses signal conditioning and power amplification to encode the coprime-coded progressive phase-offset signal to generate one of the plurality of transmit signals and transmits one of the plurality of transmit signals as a radio frequency-coded transmit signal via a transmit antenna.

2. The radar system according to claim 1, characterized in that, The Doppler disambiguation module is configured to construct a binary FIR transmitter filter for each Doppler split-multiplexing (DDM) transmitter module, having a tap delay that follows a unique coprime interval sequence associated with the transmitter module.

3. The radar system according to claim 2, characterized in that, The Doppler disambiguation module is configured to perform threshold detection on the Doppler spectrum generated from the digital signal and generate a binary detection sequence of Doppler spectrum samples that have '1' only at the entry corresponding to the detected unit.

4. A method for operating a Doppler split-multiplexing (DDM) multiple-input multiple-output (MIMO) radar system, characterized in that, include: Generates an LFM linear frequency modulated pulse signal; Multiple Doppler split-multiplexed DDM transmitter modules generate multiple Doppler split-multiplexed DDM transmit signals from the LFM linear frequency modulated pulse signal. Each of the multiple Doppler split-multiplexed DDM transmit signals has a coprime coded CPC phase offset that is different from the LFM linear frequency modulated pulse signal. At one or more receiving antennas of the receiver module, the target return signal reflected by the target from the plurality of Doppler split-multiplexed (DDM) transmitted signals is received; At the receiver module, the target return signal is mixed with the LFM linear frequency modulated pulse signal to generate an intermediate frequency signal; The intermediate frequency signal is converted into a digital signal using an analog-to-digital converter at the receiver module. Process the digital signal to detect Doppler spectral peaks in the digital signal; as well as A coprime-coded CPC decoder is applied to associate each detected Doppler spectral peak with the corresponding Doppler split-multiplexed DDM transmitter module, thereby generating multiple Doppler spectral peak detections associated with the transmitter; The generation of the plurality of Doppler split-multiplexed (DDM) transmit signals includes applying a plurality of phase shifts to the LFM linear frequency modulated (LFM) pulse signals using a plurality of corresponding multi-bit slow time phase shifters to generate a plurality of coprime-coded CPC asymptotic phase shift signals from the LFM pulse signals, each coprime-coded CPC asymptotic phase shift signal having a different coprime-coded CPC phase shifted from the LFM pulse signals.

5. The method according to claim 4, characterized in that, The coprime-coded CPC decoder includes constructing a binary FIR transmitter filter for each Doppler split-multiplexed (DDM) transmitter module, having a tap delay that follows a unique coprime interval sequence associated with the Doppler split-multiplexed (DDM) transmitter module.

6. The method according to claim 5, characterized in that, The application of the coprime-coded CPC decoder includes: Perform threshold detection on the Doppler spectrum generated from the digital signal; and Generate a binary detection sequence that has a '1' only at the entry corresponding to the detected cell.

7. A radar system, characterized in that, include: Multiple transmitter modules are configured to simultaneously transmit multiple Doppler split-multiplexed (DDM) linear frequency modulated pulse waveforms in a single transmission frame with a fixed pulse repetition frequency (PRF) rate. Each Doppler split-multiplexed DDM linear frequency modulated pulse waveform is encoded using a programmable multi-bit phase rotator with different coprime coded CPC phase offsets, such that the Doppler spectral signals from different transmitter modules can be separated in a single transmission frame when there is overlap between the spectra from different transmitter modules. At least a first receiver module is connected and configured to receive a target return signal reflected from the plurality of Doppler split-multiplexed DDM linear frequency modulated pulse waveforms through at least one target and to generate a digital signal based on the target return signal; as well as A radar control processing unit is configured to detect Doppler spectral peaks in the digital signal and apply coprime coding (CPC) decoding to associate each detected Doppler spectral peak with a corresponding transmitter module, thereby generating multiple Doppler spectral peak detections associated with the transmitter. Each transmitter module includes: A slow time phase shifter is connected to generate a coprime-coded progressive phase shift signal from a transmitted reference signal. as well as A signal encoder uses signal conditioning and power amplification to encode the coprime-coded progressive phase-shift signal to generate a transmit signal from one of the plurality of Doppler split-multiplexed (DDM) linear frequency modulated pulse waveforms, and transmits the transmit signal from one of the plurality of Doppler split-multiplexed (DDM) linear frequency modulated pulse waveforms as a radio frequency encoded transmit signal via a transmit antenna.

8. The radar system according to claim 7, characterized in that, The radar control processing unit is configured to: For each transmitter module, a binary FIR transmitter filter is constructed with a tap delay that follows a unique coprime interval sequence associated with the transmitter module; Threshold detection is performed on the Doppler spectrum generated from the digital signal, and a binary detection sequence is generated that has '1' only at the entry corresponding to the detected cell; The binary detection sequence is filtered by each binary FIR transmitter filter to associate the spectral peak with the corresponding Doppler split-multiplexed (DDM) transmitter module when the output of the corresponding filter exceeds a predefined threshold, thereby identifying the detection associated with the transmitter; and For each transmitter-associated detection, the variance of the magnitude of the Doppler spectrum sample that matches the tap delay position in the binary FIR transmitter filter is evaluated against an evaluation threshold to confirm that each transmitter-associated detection has a variance equal to or lower than the evaluation threshold.

Images