FMCW radar system with synchronized virtual antenna array

Abstract

In described examples, a frequency modulated continuous wave (FMCW) radar system includes: a first FMCW device including a first processor; and a second FMCW device including a second processor. The first and second processors receive first and second sets of FMCW signals corresponding to a field of view (FOV), respectively, and process the first and second sets of FMCW signals independently of each other to generate first and second sets of virtual antenna array signals, respectively. The second FMCW device transmits the second set of virtual antenna array signals to the first FMCW device. The first processor determines angle of arrival information relative to one or more objects in the FOV in response to the first and second sets of virtual antenna array signals.

Description

Technical Field

[0001] This application generally relates to frequency modulated continuous wave (FMCW) radar, and more particularly, to using a virtual antenna array in an FMCW system to improve the accuracy of object position determination. Background Art

[0002] An FMCW radar transmits an electromagnetic radiation (EMR) signal having a known frequency, which is modulated to vary up and down over time. The radar receives a reflected signal corresponding to the transmitted signal and uses the received signal to determine the presence, distance, angle of arrival, velocity, and direction of movement of an object within the detection range limit of the FMCW radar. The velocity and direction of movement together correspond to the speed of the detected object. Summary of the Invention

[0003] In the described example, a frequency modulated continuous wave (FMCW) radar system includes: a first FMCW device including a first processor; and a second FMCW device including a second processor. The first and second processors respectively receive first and second sets of FMCW signals corresponding to a field of view (FOV) and independently process the first and second sets of FMCW signals to respectively generate first and second sets of virtual antenna array signals. The second FMCW device transmits the second set of virtual antenna array signals to the first FMCW device. The first processor determines angle of arrival information with respect to one or more objects in the FOV in response to the first and second sets of virtual antenna array signals. Brief Description of the Drawings

[0004] Figure 1A is a graph of an example FMCW signal to be transmitted by an FMCW radar system.

[0005] Figure 1B is a graph of an example data frame of an FMCW signal to be transmitted by an FMCW radar system.

[0006] Figure 2A is a diagram of an example Doppler division multiple access (DDMA) FMCW transmission.

[0007] Figure 2B is corresponding to Figure 2A is a graph of an example received reflected FMCW signal of a DDMA FMCW transmission.

[0008] Figure 3 is for as Figure 2A shown transmitting a DDMA FMCW transmission and as Figure 2B shown receiving a reflected FMCW chirp is a functional block diagram of an example FMCW radar system.

[0009] Figure 4 Describe the process for determining distance and rate using FMCW chirps transmitted and received by an FMCW radar system Figure 3 .

[0010] Figure 5 Describe a set of two-dimensional fast Fourier transforms (FFTs) generated by applying the process Figure 4 to the DDMA FMCW signals received by the first, second, third, and fourth receivers Figure 3 .

[0011] Figure 6 is a schematic diagram of an example multi-transceiver FMCW radar system 600.

[0012] Figure 7A is a functional block diagram showing an example multi-transceiver FMCW radar system with a shared reference clock.

[0013] Figure 7B is a functional block diagram showing an example multi-transceiver FMCW radar system 718, where the first FMCW transceiver 702 and the second FMCW transceiver 704 each have a respective local reference clock for generating a reference clock signal.

[0014] Figure 8 is an example process for performing object detection using the multi-transceiver FMCW radar system of FIG. 7.

[0015] Figure 9 is a set of graphs illustrating example differences between parameters of FMCW chirps to be transmitted by the multi-transceiver FMCW radar system of FIG. 7. DETAILED DESCRIPTION

[0016] In this document, some different but closely related structures or signals have reference designators in the [digit][letter] format, such as transmitters 206a, 206b, and 206c and receivers 310a, 310b, 310c, and 310d. In some examples, these structures or signals are generally referred to in the singular form or as a group using [digit] without [letter], such as transmitter 206 and receiver 310. Additionally, the same reference designator or other reference specifier is used in the schematic diagram to indicate features that are closely related structurally and / or functionally.

[0017] In this document, references to transmitters and receivers herein refer to the corresponding transmitting or receiving antennas. In some instances, the antenna is located separately from the integrated circuit (IC) (e.g., on different parts of a printed circuit board (PCB)) and is electrically connected to the integrated circuit (IC), which contains other parts of the corresponding transmitter-related and receiver-related structures.

[0018] Figure 1A It is a graph of an example FMCW signal 100 to be transmitted by an FMCW radar system (e.g., an automotive or industrial FMCW radar system). The vertical axis corresponds to frequency, and the horizontal axis corresponds to time. The FMCW signal 100 is modulated to define an FMCW chirp 102. The duration of the FMCW chirp 102 is referred to herein as the ramp time 104. During the FMCW chirp 102, the transmitter frequency can ramp up from a base frequency F0 106 to a maximum frequency F1 108 like a sawtooth wave. The FMCW chirp 102 has a slope S = (F1 - F0) / ramp time. The slope of the FMCW signal 100 corresponds to the frequency change per unit time, e.g., ΔHz / s. Between the FMCW chirps 102 is an idle time 110. The idle time 110 is a period during which the transmitter transmitting the FMCW signal 100 is turned off.

[0019] The time from the start of one FMCW chirp 102 of the same transmitter to the start of the next FMCW chirp 102 is referred to as the pulse repetition interval (PRI) 112 of the FMCW signal 100 and is equal to the ramp time 104 plus the idle time 110 of the transmitter. Equivalently, when interleaving chirps 102 from different transmitters having similar ramp times 104 and idle times 110, the PRI 112 is equal to the number of transmitters in the FMCW radar system multiplied by the ramp time 104 of one FMCW chirp 102 plus the sum of the time from the end of an FMCW chirp 102 of another transmitter in the FMCW radar system to the start of the next sequential FMCW chirp 102 (ramp time 204 plus the inter-chirp time). In some instances, the duration of the PRI 112 defines the maximum distinguishable Doppler range. The reciprocal of the PRI 112 is the pulse repetition frequency (PRF) of the FMCW signal 100.

[0020] Fast time refers to the different time slots that make up the PRI 112 during a single FMCW chirp 102 and can depend on the rate at which the received signal is sampled. Slow time is updated after each PRI 112 and refers to the time over the course of multiple FMCW chirps 102.

[0021] Figure 1B It is a graph 114 of an example data frame 116 of the FMCW signal 100 to be transmitted by an FMCW radar system. The data frame 116 includes an acquisition period 118 and an inter-frame period 120. The acquisition period includes from various transmitters 206 (see Figure 2A)A series of FMCW chirps 102. During the inter-frame period 120, the transmitter 206 does not transmit a signal. In some instances, the inter-frame period 120 is used to save power or provide time to complete the analysis of the received reflected signals corresponding to the FMCW chirps 102 transmitted during the acquisition period 118.

[0022] Figure 2A is a diagram of an exemplary Doppler division multiple access (DDMA) FMCW transmission 200. DDMA is used herein as an example transceiver protocol; for example, other transceiver protocols such as time division multiple access (TDMA) or binary phase modulation can also be used with the methods and systems described herein. The FMCW synthesizer 202 (see Figure 3 , also referred to as the FMCW signal generator) generates the FMCW signal 100. The FMCW synthesizer 202 outputs the FMCW signal 100 to a first phase shifter (phase shifter 1) 204a, a second phase shifter (phase shifter 2) 204b, and a third phase shifter (phase shifter 3) 204c. The first phase shifter 204a outputs the phase-shifted FMCW signal 100 to the first transmitter 206a. The second phase shifter 204b outputs the phase-shifted FMCW signal 100 to the second transmitter 206b. The third phase shifter 204c outputs the phase-shifted FMCW signal 100 to the third transmitter 206c.

[0023] Shows example FMCW chirps 102 corresponding to each of the first transmitter 206a, the second transmitter 206b, and the third transmitter 206c. The FMCW synthesizer 202 and the first phase shifter 204a together generate a first set of chirps 208a. The FMCW synthesizer 202 and the second phase shifter 204b together generate a second set of chirps 208b. The FMCW synthesizer 202 and the third phase shifter 204c together generate a third set of chirps 208c.

[0024] To perform DDMA FMCW transmission, phase encoding is used to phase-shift the FMCW signal (e.g., FMCW signal 100 of FIG. 1), and the phase encoding differentiates between slow times (e.g., between chirp sets) for different transmitters. In an example, the phase encoding has sixty-four possible code settings from zero to sixty-three. Each code increment represents π / 64 radians, such that the phase-shift range is from zero radians to 2π*63 / 64 radians. Example phase-code vectors include [0 16 32 48 0 16 32 48 0] and [0 24 48 8 32 56 16 40 0]. The first phase-code vector increments by 16, and the second code vector increments by 24. The phase-shift increment corresponding to the code increment of the phase-code vector is referred to herein as the base phase shift. After eight FMCW chirps 102, the two phase-code vectors return to the same zero-valued phase offset, such that they periodically have the same value in fast time. However, the two vectors are differentiated in slow time.

[0025] In an example, the base phase shifts of the first phase shifter 206a, the second phase shifter 206b, and the third phase shifter 206c are φ1 = 0, φ2 = υ, and φ3 = 2υ, respectively, where υ is the phase shift corresponding to an integer code value. The transmitters 206a, 206b, and 206c transmit FMCW chirps that have been phase-shifted using successively increasing multiples of the respective base phase shifts. These increasing phase-shift cycles pass through the respective phase-code vectors.

[0026] The DDMA FMCW radar system can be used to implement a multiple-input multiple-output (MIMO) radar system. In a MIMO radar system having a number N of transmitters and a number M of receivers, the N signals transmitted by the transmitters are predictable and different across different transmitters. In some examples, in DDMA, the transmitted signals are differentiated by applying a unique Doppler frequency-shift sequence for each transmitter to the corresponding set of FMCW chirps 102 to be transmitted. In TDMA, the transmitted signals are differentiated by having each transmitter transmit in a time slot unique relative to the other transmitters. In binary phase modulation, each transmitter has a unique phase sequence across slow time, corresponding to a sequential phase shift of 0° or 180°, which enables signal recovery at the receiver.

[0027] A MIMO radar system that uses signals differentiated across transmitters, as described, extracts N different signals from each of the signals received by M receivers, resulting in N x M different received signals, as if the MIMO radar system had N x M different receivers. This enables improved spatial resolution of the radar system. Doppler differentiation can be used to make the N transmitted signals predictable and unique using phase-shifted vectors that are differentiated from each other in slow time. Due to the property of being able to extract N x M different received signals, the MIMO radar system as described is referred to as having an N x M element virtual antenna array.

[0028] Figure 2B corresponds to Figure 2A is a graph of an example Doppler shift of the received reflected FMCW chirp 210 of the DDMA FMCW transmission 200 corresponding to Figure 3 The horizontal axis corresponds to the Doppler shift frequency and the vertical axis corresponds to the amplitude. The Doppler shift refers to the frequency change of the signal received by an FMCW radar system (e.g.,

[0029] The Doppler shift of the first received signal (signal Tx1a) 212a corresponds to the FMCW chirp 102 of the first set of chirps 208a transmitted by the first transmitter 206a. The Doppler shift of the second received signal (signal Tx2a) 214a corresponds to the FMCW chirp 102 of the second set of chirps 208b transmitted by the second transmitter 206b. The Doppler shift of the third received signal (signal Tx3a) 216a corresponds to the third set of chirps 208c transmitted by the third transmitter 206c. Signals Tx1a 212a, signal Tx2a 214a, and signal Tx3a 216a are shown grouped together, separated in frequency by relatively small increments corresponding to the phase separation of the first set of chirps 208a, the second set of chirps 208b, and the third set of chirps 208c. Thus, signals Tx1a 212a, signal Tx2a 214a, and signal Tx3a 216a correspond to the first detected object.

[0030] Similarly, the Doppler shift of the fourth received signal (signal Tx1b) 212b corresponds to the FMCW chirp 102 of the first set of chirps 208a transmitted by the first transmitter 206a. The Doppler shift of the fifth received signal (signal Tx2b) 214b corresponds to the FMCW chirp 102 of the second set of chirps 208b transmitted by the second transmitter 206b. The Doppler shift of the sixth received signal (signal Tx3b) 216b corresponds to the third set of chirps 208c transmitted by the third transmitter 206c. The signals Tx1b 212b, signal Tx2b 214b, and signal Tx3b 216b are shown grouped together and separated in frequency by relatively small increments corresponding to the phase separation of the first set of chirps 208a, the second set of chirps 208b, and the third set of chirps 208c. Thus, the signals Tx1b 212b, signal Tx2b 214b, and signal Tx3b 216b correspond to the second detected object.

[0031] Figure 3 is for as Figure 2A shown to transmit the DDMA FMCW transmission 200 and as Figure 2B shown to receive the reflected FMCW chirps 210 of the example FMCW radar system 300. In some instances, the DDMA FMCW radar system or an FMCW radar system using another type of transceiver protocol (e.g., TDMA or binary phase modulation) uses different functional blocks. In some instances, the FMCW radar system 300 is configured to use millimeter wave sensing or sub-terahertz (sub-THz) sensing. In some instances, the FMCW radar system 300 using millimeter wave sensing transmits the FMCW chirps 102 in the 60 gigahertz or 77 gigahertz frequency bands. In some instances, the FMCW radar system 300 using sub-THz sensing transmits the FMCW chirps 102 in a frequency band of 140 gigahertz (GHz) or higher.

[0032] The FMCW radar system 300 includes an FMCW synthesizer 202, a digital signal processor (DSP) 302, a transmitter side 304, a receiver side 306, a temperature sensor 319, and a memory 320. The transmitter side 304 of the FMCW radar system 300 includes a first phase shifter (phase shifter 1) 204a, a second phase shifter (phase shifter 2) 204b, and a third phase shifter (phase shifter 3) 204c; a first power amplifier (PA1) 308a, a second power amplifier (PA2) 308b, and a third power amplifier (PA3) 308c; and a first transmitter (TX1) 206a, a second transmitter (TX2) 206b, and a third transmitter (TX3) 206c.

[0033] The receiver side 306 of the FMCW radar system 300 includes a first receiver (RX1) 310a, a second receiver (RX2) 310b, a third receiver (RX3) 310c, and a fourth receiver (RX4) 310d; a first low-noise amplifier (LNA1) 312a, a second low-noise amplifier (LNA2) 312b, a third low-noise amplifier (LNA3) 312c, and a fourth low-noise amplifier (LNA4) 312d; a first mixer 314a, a second mixer 314b, a third mixer 314c, and a fourth mixer 314d; a first band-pass filter (BPF) and variable gain amplifier (VGA) circuit (BPF / VGA 1) 316a, a second BPF and VGA circuit (BPF / VGA 2) 316b, a third BPF and VGA circuit (BPF / VGA 3) 316c, and a fourth BPF and VGA circuit (BPF / VGA 4) 316d; a first analog-to-digital converter (ADC) circuit (ADC 1) 318a, a second ADC circuit (ADC 2) 318b, a third ADC circuit (ADC 3) 318c, and a fourth ADC circuit (ADC 4) 318d.

[0034] The FMCW synthesizer 202 generates an FMCW chirp 102 to be transmitted, for example, for object detection and range, angle, and velocity determination. The FMCW synthesizer 202 outputs the FMCW chirp 102 to corresponding first inputs of a first phase shifter 204a, a second phase shifter 204b, and a third phase shifter 204c, and also outputs to corresponding first inputs of a first mixer 314a, a second mixer 314b, a third mixer 314c, and a fourth mixer 314d. The first phase shifter 204a, the second phase shifter 204b, and the third phase shifter 204c phase-shift the FMCW chirp 102 using corresponding phase-shift code vectors, as described with respect to Figure 2A as described.

[0035] The first phase shifter 204a, the second phase shifter 204b, and the third phase shifter 204c respectively output the FMCW chirp 102 to a first power amplifier PA1 308a, a second power amplifier PA2 308b, and a third power amplifier PA3 308c. The first power amplifier PA1 308a, the second power amplifier PA2 308b, and the third power amplifier PA3 308c amplify the corresponding phase-shifted FMCW chirp signals and output the amplified signals to a first transmitter 206a, a second transmitter 206b, and a third transmitter 206c respectively. The first transmitter 206a, the second transmitter 206b, and the third transmitter 206c transmit the amplified phase-shifted FMCW chirps. In some instances, the transmitted signals are reflected by an object 322 (object 322 in range) within the detection and range, angle, and velocity determination range of the FMCW radar system 300.

[0036] The reflected signals are received by a first receiver 310a, a second receiver 310b, a third receiver 310c, and a fourth receiver 310d. The first receiver 310a, the second receiver 310b, the third receiver 310c, and the fourth receiver 310d output the received signals to a first low-noise amplifier LNA1 312a, a second low-noise amplifier LNA2 312b, a third low-noise amplifier LNA3 312c, and a fourth low-noise amplifier LNA4 312d that amplify the received signals, respectively. The first low-noise amplifier LNA1 312a, the second low-noise amplifier LNA1 312b, the third low-noise amplifier LNA1 312c, and the fourth low-noise amplifier LNA4 312d output the amplified signals to a second input of a first mixer 314a, a second mixer 314b, a third mixer 314c, and a fourth mixer 314d, respectively. The first mixer 314a, the second mixer 314b, the third mixer 314c, and the fourth mixer 314d output the mixed signals to a first BPF / VGA circuit 316a, a second BPF / VGA circuit 316b, a third BPF / VGA circuit 316c, and a fourth BPF / VGA circuit 316d, respectively, which filter and amplify the mixed signals. The first BPF / VGA circuit 316a, the second BPF / VGA circuit 316b, the third BPF / VGA circuit 316c, and the fourth BPF / VGA circuit 316d output the resulting cleaned signals to a first ADC circuit 318a, a second ADC circuit 318b, a third ADC circuit 318c, and a fourth ADC circuit 318d, respectively, which sample the cleaned mixed signals to generate corresponding data sets consisting of digital samples. The first ADC circuit 318a, the second ADC circuit 318b, the third ADC circuit 318c, and the fourth ADC circuit 318d output the digital samples to the DSP 302 for analysis.

[0037] The DSP 302 uses the digital samples to determine the presence, distance, angle, and velocity of an object 322 in range. In this context, an object in range refers to an object within the shared field of view (FOV) of the FMCW transmitter 206 and the corresponding FMCW receiver 310 and within the designed range of the corresponding FMCW radar system (e.g., Figure 3 the FMCW radar system 300) where the received reflected FMCW signals can be used to detect the object.

[0038] For example, the presence of an object can be determined based on a signal amplitude greater than a threshold. The distance can be determined by multiplying the unique range frequency corresponding to the round-trip delay of the signal by the slope of the FMCW chirp 102. The velocity can be determined by the phase change of the unique range frequency, which appears as a unique Doppler frequency, over multiple chirps. The angle can be determined by the phase change of a particular received chirp across different receivers, which is caused by the difference in time of flight across different receivers. Further discussion of these determinations is provided with respect to Figure 4 and 5 these determinations.

[0039] Figure 4 Describe a process 400 for determining range and velocity using the FMCW chirps 102 transmitted and received by the FMCW radar system 300 of Figure 3 For step 402, the horizontal axis indicates time and the vertical axis indicates frequency. For step 404, the horizontal axis indicates time and the vertical axis indicates amplitude. The IF signal is the product of mixing the received signal and the transmitted signal. In step 402, the FMCW signal 100 is transmitted and the received FMCW signal 406 is received. Each FMCW chirp 102 is transmitted and received in fast time. The FMCW signal 100 is transmitted and the received FMCW signal 406 is received in slow time.

[0040] The amount of time for the transmitted signal to reach the object 322 at the range is equal to d. The time for the reflected signal to return from the object 322 at the range and be received by the first receiver 310a, the second receiver 310b, the third receiver 310c, and the fourth receiver 310d is also equal to d. Thus, the time of flight of the FMCW chirp 102 reflected by the object 322 at the range is 2d. In some instances, the value of d varies in response to the different positions of different transmitters in the transmitter 306 and / or the different positions of the receivers 410. This varying value of d appears as a phase change in the signals received by the different receivers 410 for performing angle estimation. Additionally, as discussed with respect to Figure 2B the received FMCW chirp 102 is Doppler shifted with respect to the corresponding transmitted FMCW chirp 102, which depends on the motion of the FMCW radar system 300 with respect to the object 322 at the range (from which the received FMCW chirp 102 is reflected), and depends on the phase shift applied by the corresponding phase shifter in the first phase shifter 204a, the second phase shifter 204b, or the third phase shifter 204c.

[0041] In step 404, the first mixer 314a, the second mixer 314b, the third mixer 314c, and the fourth mixer 314d mix (e.g., multiply) the respective received signals with the FMCW signal 100 generated by the FMCW synthesizer 202 to produce intermediate frequency (IF) signals 408. The frequency of the IF signals is linearly proportional to the time of flight 2d of the corresponding FMCW chirp 102. As described with respect to Figure 3 the first ADC 318a, the second ADC 318b, the third ADC 318c, and the fourth ADC 318d sample the IF signals after they are cleaned and amplified, and provide the resulting digital samples to the DSP 302 for analysis. In step 410, the DSP 302 performs a fast Fourier transform (FFT) on the set of digital samples in fast time. This means determining the FFT for the set of samples of the IF signals (the received signals mixed with the transmitted signals) such that the set of samples is aligned with the respective PRI 112. This results in a series of one-dimensional range FFTs 412 in time sequence.

[0042] The range FFTs 412 are divided into frequency bins 414. Each frequency bin 414 covers a separate Doppler shift frequency range and has an index indicating the distance to the object and a value indicating the return signal strength associated with the respective distance. The number of frequency bins 414 in the respective range FFTs 412 corresponds to the frequency resolution of the FMCW radar system 300. The frequency resolution of the FMCW radar system 300 corresponds to the range and rate resolution of the FMCW radar system 300.

[0043] In the illustrated example, there are eight frequency bins 414 in each range FFT 412. In some examples, the range FFTs 412 contain hundreds of frequency bins. If there is an object 322 in range that reflects the transmitted FMCW chirp 102 over a period of time, there will be amplitude spikes 416. The amplitude spikes 416 are shown as shaded boxes in the frequency bins 414 of the range FFT 412 in Figure 4 The amplitude spikes 416 correspond to the distance of the object 332 in range from the receiver (the first receiver 310a, the second receiver 310b, the third receiver 310c, or the fourth receiver 310d) that received the FMCW signal 100 being analyzed. The range FFT 412 with this amplitude spike 416 corresponds to the intermediate frequency indicating the presence of the object 322 in range.

[0044] In step 418, the DSP 302 performs an FFT on the one-dimensional range FFT in slow time. Thus, the DSP 302 performs an FFT on the time-ordered set of one-dimensional range FFTs 412 to produce a two-dimensional range-Doppler FFT 420. The range-Doppler FFT 420 contains a set of bins, each bin having (1) an index representing a combination of range and rate, and (2) a value indicating the return signal strength associated with the corresponding range and rate. The range-Doppler FFT 420 covers multiple PRIs 112 determined in response to the designed rate resolution; in some instances, this corresponds to one data frame 116. The vertical dimension of the range-Doppler FFT 420 corresponding to fast time (relative to the individual PRIs 112 of the corresponding transmitter in the transmitter 206) is divided into frequency bins 414 indicating range. The vertical dimension of the range-Doppler FFT 420 is also referred to as the range domain of the range-Doppler FFT 420. The horizontal dimension of the range-Doppler FFT 420 corresponding to slow time (spanning the selected number of PRIs 112) is divided into frequency bins 414 indicating Doppler shift. The horizontal dimension of the range-Doppler FFT 420 is also referred to as the Doppler domain of the range-Doppler FFT 420. In some instances, the selected number of PRIs 112 covers several tens of milliseconds.

[0045] The amplitude spike 422 (dark box) in the range-Doppler FFT 420 indicates the presence of an object 322 in range. The vertical coordinate of the particular frequency bin 414 in which the amplitude spike 422 is located indicates the range of the object 322 in range from the FMCW radar system 300. The horizontal coordinate of the particular frequency bin 414 in which the amplitude spike 422 is located provides Doppler shift information. The Doppler shift information represented by the amplitude spike 422 in the range-Doppler FFT 420 can be used to determine the velocity of the object 322 in range relative to the FMCW radar system 300. In an instance, the determined velocity is the average velocity over the selected number of PRIs 112 used to generate the range-Doppler FFT 420.

[0046] Figure 5 Illustrated by applying Figure 4 the process 400 to that by Figure 3A set of 500 range-Doppler FFTs 420 is generated from the DDMAFMCW signals received by the first receiver 310a, the second receiver 310b, the third receiver 310c, and the fourth receiver 310d. As described above, the differential phase shift vectors applied to the first phase shifter 204a, the second phase shifter 204b, and the third phase shifter 204c using slow time enable the FMCW signal 100 transmitted by the number N of transmitters and received by the number M of receivers to be regarded as N×M individual received signals. For each of the M receivers, an object 322 in range will appear as N different peaks in the range FFT 412. This increases the spatial resolution of the FMCW radar system 300.

[0047] The FMCW radar system 300 has three transmitters and four receivers. Thus, applying the process 400 to the FMCW radar system 300 results in twelve received signals, which can also be regarded as twelve objects to be resolved. A disambiguation step (also referred to as transmitter decoding) is performed to distinguish the twelve objects, and then the corresponding range FFTs 412 are processed to generate twelve range-Doppler FFTs 420. Different objects among the distinguished objects correspond to different combinations of the first transmitter 206a, the second transmitter 206b, or the third transmitter 206c and the first receiver 310a, the second receiver 310b, the third receiver 310c, or the fourth receiver 310d, such that different range-Doppler FFTs in the range-Doppler FFTs 420 correspond to different transmitter-receiver combinations. The range-Doppler FFTs 420 are identified according to their corresponding transmitters and receivers. For example, the range-Doppler FFT 420 corresponding to the first transmitter (TX1) 206a and the third receiver (RX3) 310c is identified as TX1-RX3, and the range-Doppler FFT 420 corresponding to the third transmitter (TX3) 206c and the second receiver (RX2) 310b is identified as TX3-RX2.

[0048] As described above, by using the range-Doppler FFTs 420 corresponding to multiple different receivers, the angle (angle of arrival) of the orientation of the object 322 in range relative to the FMCW radar system 300 can be determined. In some instances, this is done by performing an FFT (referred to as an angle FFT) across the range-Doppler FFTs 420. For example, two receivers can be used to determine the angle in a single plane, which can be combined with range to generate the two-dimensional position of the object 322 in range. For example, two receivers can be used to determine the range and azimuth angle of the object 322 in range. Similarly, three receivers can be used to determine the angles in multiple planes, which can be combined with range to determine the three-dimensional position of the object 322 in range. For example, three receivers can be used to determine the range, azimuth angle, and elevation angle of the object 322 in range.

[0049] The accuracy of determining the angle information of an object 322 in a range is limited by the number of antennas used to receive reflected signals. In a MIMO radar system, this limitation corresponds to the number of virtual antennas in a virtual antenna array.

[0050] Figure 6 FIG. 6 is a diagram of an exemplary multi-transceiver FMCW radar system 600. The multi-transceiver FMCW radar system 600 includes a first FMCW transceiver 602 and a second FMCW transceiver 604 connected by a communication interface 606. In this document, an "FMCW transceiver" or "FMCW device" is used to refer to a device that includes a structure (e.g., circuitry) for generating and transmitting FMCW chirps, receiving FMCW signals, and processing FMCW signals. In some instances, the first FMCW transceiver 602 is physically separated from the second FMCW transceiver 604. As further explained with respect to FIG. 7, the first FMCW transceiver 602 and the second FMCW transceiver 604 are MIMO radar systems. Figure 3 The FMCW radar system 300 of FIG. 8 provides an exemplary FMCW device that can be used to implement the first FMCW transceiver 602 and / or the second FMCW transceiver 604.

[0051] The first FMCW transceiver 602 has a first FOV 608, and the second FMCW radar system 604 has a second FOV 610. The first FOV 608 and the second FOV 610 overlap. The first FMCW radar system 602 and the second FMCW radar system 604 use the communication interface 606 to synchronize FMCW chirp transmissions and share information about received signals to improve angle determination with respect to objects located in the overlapping region of the first FOV 608 and the second FOV 610, which is referred to herein as the shared FOV of the first FMCW transceiver 602 and the second FMCW transceiver 604.

[0052] Figure 7Ais a functional block diagram showing an example multi-transceiver FMCW radar system 700 having a shared reference clock 712. The multi-transceiver FMCW radar system 700 also includes a first FMCW transceiver 702 and a second FMCW transceiver 704 connected by a data communication line or bus 706. The first FMCW transceiver 702 includes a first local oscillator 714, and the second FMCW transceiver 704 includes a second local oscillator 716. The first local oscillator 714 and the second local oscillator 716 generate respective relatively high-frequency reference signals, such as 20 GHz signals. These high-frequency reference signals are used to generate respective FMCW chirps 102. The shared reference clock 712 generates a relatively low-frequency reference signal, such as a 40 MHz signal. Also, for example, a synchronization pulse is transmitted from the shared reference clock 712 or from the first FMCW transceiver 702 to the second FMCW transceiver 704. The shared reference clock signal and the synchronization pulse are used to achieve automatic synchronization of FMCW chirp parameters across the first FMCW transceiver 702 and the second FMCW transceiver 704, the FMCW chirp parameters including start frequency, FMCW chirp slope, ADC start time, and inter-chirp time.

[0053] The first FMCW transceiver 702 includes a number P of virtual antennas 708, as described with respect to Figure 2A as described. The second FMCW transceiver 704 includes a number R of virtual antennas 710. In some instances, the first FMCW transceiver 702 may be modeled as a pilot circuit of the multi-transceiver FMCW radar system 700, and the second FMCW transceiver 704 may be modeled as a follower circuit of the multi-transceiver FMCW radar system 700. As described herein with respect to the virtual antenna array synchronization described in accordance with Figure 7A and 7B enables the use of the first FMCW transceiver 702 and the second FMCW transceiver 704 to accurately determine angle information corresponding to a virtual antenna array having P + R virtual antennas. Virtual antenna array synchronization is performed using FMCW chirp parameter synchronization.

[0054] Figure 7B is a functional block diagram showing an example multi-transceiver FMCW radar system 718, where the first FMCW transceiver 702 and the second FMCW transceiver 704 each have a respective local reference clock that generates a reference clock signal. The first FMCW transceiver 702 includes a first reference clock 720, and the second FMCW transceiver 704 includes a second reference clock 722. The first FMCW radar system 702 sends an Ethernet-PTP timestamp and a frame synchronization pulse to the second FMCW transceiver 704 to achieve virtual antenna synchronization of the example FMCW radar system 714 in accordance with Figure 7B as described below. Virtual antenna synchronization of the example FMCW radar system 714 in accordance with Figure 7A and7B Synchronization of the virtual antenna arrays of exemplary FMCW radar systems 700 and 714.

[0055] The first FMCW transceiver 702 and the second FMCW transceiver 704 independently transmit and receive FMCW chirp signals from each other. The first FMCW transceiver 702 processes its corresponding received FMCW chirp signal corresponding to its own transmission to generate a virtual antenna array signal independently of the second FMCW transceiver 704. Similarly, the second FMCW transceiver 704 processes its corresponding received FMCW chirp signal corresponding to its own transmission to generate a virtual antenna array signal independently of the first FMCW transceiver 702. In some instances, processing the FMCW chirp signal to generate the virtual array signal includes performing range and Doppler FFTs, corresponding to steps 410 and 418 (as described above).

[0056] Once each device has generated a signal corresponding to its own virtual antenna, the second FMCW transceiver 704 sends its virtual antenna information (e.g., its corresponding range-Doppler FFT 420 for data frame 116) to the first FMCW transceiver 702. (In some instances, the range FFT 412 is also provided.) The first FMCW transceiver performs angle estimation using the combined virtual antennas of the transceivers of the multi-transceiver FMCW radar system 700. This is achieved by synchronizing the first FMCW transceiver 702 and the second FMCW transceiver 704 and by avoiding interference between the first FMCW transceiver 702 and the second FMCW transceiver 704. This process is further described with respect to Figure 8 and 9 further described.

[0057] The first FMCW transceiver 702 and the second FMCW transceiver 704 each use a reference clock signal from a shared reference clock 712 or from a corresponding first reference clock 720 or second reference clock 722 to provide root timing for timing and FMCW signal generation. Different frequency clocks for different functions (or groups of functions) are generated from the reference clock signal using phase-locked loops (PLLs), multipliers, and dividers, and other frequency modification circuitry. Thus, in some instances, the corresponding relatively high-frequency local oscillators 714 and 716 are derived from the reference clock signal within the respective FMCW transceivers 702 or 704 (from a shared or transceiver-specific reference clock). In some instances, a shared reference clock signal or individual reference clock signals are used to generate clock signals to control and synchronize functions such as the FMCW chirp 102 start frequency, inter-frame and inter-chirp timing, and analog-to-digital converter (ADC) start timing.

[0058] In some instances, it is not necessary to transmit a local oscillator signal between the first FMCW transceiver 702 and the second FMCW transceiver 704. In some instances (e.g., Figure 7A the example FMCW radar system 700), a low-frequency reference clock signal, such as from a shared reference clock 712 or from a reference clock 720 of the first FMCW transceiver 702, is transmitted instead of a local oscillator signal. As described above, a frame synchronization pulse is also transmitted from the first FMCW transceiver 702 to the second FMCW transceiver 704. The frame synchronization pulse is used to synchronize the start of the data frame 116 in the second FMCW transceiver 704 with the start of the corresponding data frame 116 in the first FMCW transceiver 702. In some instances, synchronizing the start of the data frames 116 means aligning them such that one starts with a fixed offset relative to the other. This fixed offset is used to achieve a time differentiation between the FMCW signal of the first FMCW transceiver 702 and the FMCW signal of the second FMCW transceiver 704, as further described below.

[0059] The relatively low-frequency reference clock signal can be used to synchronize the transmission and processing of the FMCW chirps 102 of the first FMCW transceiver 702 and the second FMCW transceiver 704, thereby avoiding phase offset errors, since each of the first FMCW transceiver 702 and the second FMCW transceiver 704 independently processes the signals it receives corresponding to the FMCW chirps 102 it transmits to generate its own respective virtual antenna array signals. Upconverting the low-frequency reference clock signal to a high-frequency local oscillator can result in phase noise. Phase offset between the FMCW chirps 102 transmitted by the first FMCW transceiver 702 and the FMCW chirps 102 transmitted by the second FMCW transceiver 704 can also result in phase noise. Mixing of the transmitted and received signals by the mixers of the respective transceivers 702 and 704 cancels out the described phase noise, since each of the transceivers 702 and 704 only processes the received reflected signals corresponding to the FMCW chirps 102 transmitted by the respective transceiver 702 or 704 (to generate the range FFT 412 and the range-Doppler FFT 420, i.e., the virtual antenna array signals).

[0060] In other words, the first FMCW transceiver 702 transmits a signal, receives the signal corresponding to its own transmitted signal, and processes those corresponding signals independently of other signals it may receive to generate an IF signal and a virtual antenna array signal. The first FMCW transceiver 702 generates the FMCW chirps 102 to be transmitted using the local oscillator of the first FMCW transceiver 702. The second FMCW transmitter 704 similarly independently generates its own IF and virtual antenna array signals using its own local oscillator. This independent activity enables relaxed synchronization timing requirements.

[0061] Independent processing of the first FMCW transceiver 702 and the second FMCW transceiver 704 can be maintained by differentiating the transmitted signals of the first FMCW transceiver 702 from the transmitted signals of the second FMCW transceiver 704. The differentiation prevents interference between different sets of signals, enabling the transceivers to separate the received (reflected) signals corresponding to the respectively transmitted signals from the received signals corresponding to the transmissions of other transceivers. The differentiation between the sets of transmitted signals can be performed, for example, by staggering the transmissions of the first FMCW transceiver 702 and the second FMCW transceiver 704 in time to prevent overlap in the intermediate frequency (IF) domain (see step 404). In some instances, this time differentiation can be adjusted in post-processing using Doppler-based phase compensation (also referred to herein as Doppler compensation).

[0062] In some instances, a time interval of a few microseconds is sufficient to achieve the differentiation. In some instances, Doppler frequency division multiplexing or another signal differentiation technique can be used to perform the differentiation.

[0063] In some instances, Doppler compensation is used to adjust the time differentiation as follows. Assume C represents a complex number corresponding to a particular virtual antenna (e.g., range bin 414 of range-Doppler cell 412). Doppler compensation is performed by multiplying C by Here, λ is the wavelength corresponding to the center frequency of the FMCW chirp 102 signal, T I-F is the difference between the start time of the data frame 116 of the first transceiver 702 and the start time of the data frame 116 of the second transceiver 704, and v is the relative velocity of the target corresponding to range-Doppler cell 422. In some instances, Doppler compensation is performed only on the virtual antennas 710 of the second FMCW transceiver 710. Doppler compensation is used to adjust the phase shift in the virtual antennas 710 of the second FMCW transceiver 704 caused by the motion of the object 322 in range during the time gap between the respective start times of the data frames 116 of the first FMCW transceiver 702 and the second FMCW transceiver 704.

[0064] In some examples, timestamps or other start and end time signals generated relative to a reference clock signal are used with synchronization pulses to synchronize the functions of the first FMCW transceiver 702 and the second FMCW transceiver 704, such as the start frequency of the FMCW chirp 102, inter-frame and inter-chirp timing, and ADC start timing. In some examples, these signals are transmitted from the first FMCW transceiver 702 to the second FMCW transceiver 704. In some examples, these signals are transmitted from an external processor 724 (other than a processor within the first FMCW transceiver 702 or the second FMCW transceiver 704) to the first FMCW transceiver 702 and the second FMCW transceiver 704. In examples using the external processor 724, both the first FMCW transceiver 702 and the second FMCW transceiver 704 perform synchronization, as described below with respect to Figure 8 and 9 as described with respect to the second FMCW transceiver 704. In some examples, the first FMCW transceiver 702 and the second FMCW transceiver 704 transmit independently determined virtual antenna array signal information to the external processor 724 to determine angle information. The dashed line indicates that the external processor 724 corresponds to an alternative example implementation.

[0065] Figure 8 is an example process 800 for performing object detection using the multi-transceiver FMCW radar system 700 of FIG. 7. In step 802, the first MIMO FMCW radar transceiver 702 generates a first set of FMCW chirps independently of the second MIMO FMCW radar transceiver 704. In step 804, the second MIMO FMCW radar transceiver 704 generates a second set of FMCW chirps independently of the first MIMO FMCW radar transceiver 702. In step 806, the first MIMO FMCW radar transceiver 702 transmits the first FMCW chirps into the FOV independently of the second MIMO FMCW radar transceiver 704. In step 808, the second MIMO FMCW radar transceiver 704 transmits the second set of FMCW chirps into the FOV independently of the second MIMO FMCW radar transceiver 704.

[0066] In step 810, the first MIMO FMCW radar transceiver 702 receives signals from its FOV to generate a first received signal independently of the second MIMO FMCW radar transceiver 704. In step 812, the second MIMO FMCW radar transceiver 704 receives signals from its FOV to generate a second received signal independently of the first MIMO FMCW radar transceiver 702. In step 814, the first MIMO FMCW radar transceiver 702 performs range and Doppler FFTs on the first received signal to generate a first set of virtual antenna array signals independently of the second MIMO FMCW radar transceiver 704. In step 816, the second MIMO FMCW radar transceiver 704 performs range and Doppler FFTs on the second received signal to generate a second set of virtual antenna array signals independently of the first MIMO FMCW radar transceiver 704.

[0067] In step 818, the second MIMO FMCW radar transceiver 704 transmits the second set of virtual antenna array signals to the first MIMO FMCW radar transceiver 702. In some instances, the transmitted virtual array signals correspond to the range-Doppler spectral information of the second MIMO FMCW radar transceiver 704, such as range-Doppler FFT information. In some instances, the transmitted virtual array signals correspond to the range-Doppler FFT 420 of the corresponding data frame 116 or group of data frames 116. In step 820, the first MIMO FMCW radar transceiver 702 determines the angle-of-arrival information with respect to one or more objects in the FOV in response to the first and second sets of virtual antenna array signals. In some instances, determining the angle-of-arrival information in response to the first and second sets of virtual antenna array signals corresponds to performing angle spectral estimation using the two sets of virtual antenna array signals, such as angle FFT. Thus, determining the angle-of-arrival information corresponds to performing an angle FFT across a data set amplified by including range-Doppler FFT 420 or other range-Doppler spectral information from both the first MIMO FMCW radar transceiver 702 and the second MIMO FMCW radar transceiver 704. In step 822, the first MIMO FMCW radar transceiver 702 uses the determined range, Doppler, and angle-of-arrival information to detect objects and determine the corresponding point cloud (points in space determined to correspond to the positions of the reflected FMCW chirps).

[0068] In some instances, the first MIMO FMCW radar transceiver 702 independently determines the angle of an object 322 in a distance that is not within the shared FOV with respect to the first MIMO FMCW radar transceiver 702. In some instances, the second MIMO FMCW radar transceiver 704 independently determines the angle of an object 322 in a distance that is not within the shared FOV with respect to the second MIMO FMCW radar transceiver 704.

[0069] In some instances, the first MIMO FMCW radar transceiver 702 and the second MIMO FMCW radar transceiver 704 each use a reference clock signal and a frame synchronization signal provided by the shared reference clock 712. The use of the shared reference clock 712 enables steps 802 to 816 to be performed independently of each other by the first MIMO FMCW radar transceiver 702 and the second MIMO FMCW radar transceiver 704.

[0070] In some instances (as further described with respect to Figure 9 the first MIMO FMCW radar transceiver 702 sends a start time, an end time, and a frame synchronization signal to the second MIMO FMCW radar transceiver before step 804. This enables steps 802 to 816 to be performed independently of each other by the first MIMO FMCW radar transceiver 702 and the second MIMO FMCW radar transceiver 704. In such instances, the independent execution of the second MIMO FMCW radar transceiver 704 is subsequent and in response to compensating for the frequency offset between the first reference clock 720 and the second reference clock 722 at the start of the corresponding data frame 116,

[0071] Figure 9 A set of graphs 900 illustrates example differences between parameters of FMCW chirps to be transmitted by the multi-transceiver FMCW radar system 700 of FIG. 7. The graphs 900 include a first graph 902 corresponding to a first FMCW chirp 906 to be transmitted by the first FMCW transceiver 702 and a second graph 904 corresponding to a second FMCW chirp 908 to be transmitted by the second FMCW transceiver 704. The vertical axis indicates frequency and the horizontal axis indicates time. The first FMCW chirp 906 and the second FMCW chirp 908 are numbered with subscripts starting from 1, depending on the order in which they will be transmitted. For example, the first graph 902 shows the first FMCW chirps 9061, 9062, and 9063, and the second graph 904 shows the second FMCW chirps 9081, 9082, and 9083. In the example, the first FMCW chirp 906 and the second FMCW chirp 908 have a designed start frequency of 77 GHz, a chirp duration of 20 microseconds, and a data frame duration of 15 milliseconds.

[0072] There is an inter-frame time T I-F , which corresponds to the difference between the start time of the data frame 910 of the first FMCW chirp 906 and the start time of the data frame 912 of the second FMCW chirp 908. The first FMCW chirp 906 has a start frequency f 0A and the second FMCW chirp 908 has a start frequency f 0B . The first FMCW chirp 906 has a slope S A and the second FMCW chirp 908 has a slope S B . The first FMCW chirp 906 has an inter-chirp time T IC-A and the second FMCW chirp 908 has an inter-chirp time T IC-B .

[0073] Methods for reducing the differences between the FMCW chirps of the FMCW chirps 906 and 908 of the first FMCW transceiver 702 and the second FMCW transceiver 704 and the data frames 910 and 912, and the phase errors introduced by sampling the differences of the corresponding ADC circuits 318 of the first FMCW transceiver 702 and the second FMCW transceiver 704 are described below. These methods include methods for reducing T I-F , and methods for reducing the difference between f 0A and f 0B , between S A and S B , and between T IC-A and T IC-B . Adjustments to the sampling parameters of the corresponding ADC circuits 318 are also described.

[0074] Transceiver timing synchronization is used to align the frames of the first FMCW transceiver 702 and the second FMCW transceiver 704. In other words, transceiver timing synchronization is performed to align a set of FMCW chirps 908 of the second FMCW transceiver 704 with respect to a corresponding set of FMCW chirps 906 of the first FMCW transceiver 702 in slow time, such that T I-F is controlled. Transceiver timing synchronization aligns the set of FMCW chirps 908 of the second FMCW transceiver 704 to start simultaneously with the corresponding set of FMCW chirps 906 of the first FMCW transceiver 702, or to start with a specified delay with respect to the corresponding set of FMCW chirps 906 of the first FMCW transceiver 702.

[0075] In some instances, a non-zero T I-F is specified (e.g., a T I-F)to prevent the transmitted signals from the first FMCW transceiver 702 and the second FMCW transceiver 704 from interfering with each other. Data frame timing synchronization enables accurate angle estimation in response to the cumulative number of virtual antennas using both the virtual antennas 708 and 710 of the first FMCW transceiver 702 and the second FMCW transceiver 704. Data frame timing synchronization is performed using frame synchronization pulses.

[0076] In some instances, to perform transceiver timing synchronization, Doppler-based phase compensation is used to correct for the programmed non-zero value T I-F 's effect. The Doppler-based phase compensation corrects the phase of the signals received by the respective virtual antennas 710 of the second FMCW transceiver 704 relative to the virtual antenna 708 of the first FMCW transceiver 702 based on the Doppler of the respective signals. The uncompensated data frame start timing error in T I-F that can be caused by, for example, the inter-frame variation of the delay of the frame synchronization pulse is referred to herein as δ synch .

[0077] This uncompensated frame start timing error δ synch can cause an error φ err in the phase of the signal virtual antenna, which is given by φ err = 4πvδ synch / λ, where v is the target rate (the rate corresponding to the amplitude spike 422 in the range-Doppler FFT 420), and λ is the average wavelength of the FMCW chirp 102 (e.g., 77.5 GHz for the FMCW chirp 102 with F0 = 77 GHz and F1 = 78 GHz). Thus, in some instances, the synchronization as described herein can tolerate a frame start timing error δ synch of several hundred nanoseconds (ns). For example, given an object 322 in a given distance with a maximum rate of 100 kilometers per hour, and the budget error contribution of δ synch is less than 1°, the δ synch across the device is approximately 200 ns or less. (In some instances, the angular error contribution described herein corresponds to the phase error in the frequency bin of the corresponding range-Doppler FFT 420 that contains the amplitude spike 422.)

[0078] As described above, an example method of synchronizing multiple FMCW transceivers to perform object detection uses a shared reference signal, such as a 40 MHz clock. The first FMCW transceiver 702 and the second FMCW transceiver 704 each use the shared reference signal to internally generate chirp signals, ADC timing signals, and inter-chirp timing signals. The use of the shared reference signal enables automatic synchronization of certain parameters such as start frequency, chirp slope, ADC start time, and inter-chirp time.

[0079] An alternative method for timing synchronization uses the Precision Time Protocol (PTP) supported in some Ethernet implementations. In some instances, this method is used when the use of a shared reference signal is unavailable or an Ethernet connection is available. In some instances, another communication protocol is used that similarly implements precise timestamps or otherwise implements precise timing determination of a start time and an end time, with a specified number of clock cycles generated by a reference clock between the start time and the end time. In some instances, using Ethernet-PTP (or the like) enables avoidance of using a shared reference clock. In some instances, using Ethernet-PTP for synchronization enables synchronization to be achieved when transceivers are not co-located (e.g., in a large baseline array) or when Ethernet is used for data transfer (e.g., in a satellite radar architecture). Synchronization using Ethernet-PTP (or the like) can be performed as follows.

[0080] In an instance where Ethernet-PTP is used for timing synchronization, a first FMCW transceiver 702 and a second FMCW transceiver 704 each use an independent reference clock, such as a 40 MHz clock. In some instances, these independent reference clocks, although operating at the same nominal frequency, have different instantaneous drifts in frequency. The drift between the reference clock of the first FMCW transceiver 702 and the reference clock of the second FMCW transceiver 704 is determined as follows. This drift is referred to as alpha (read as alpha) of the reference clock 722 of the second FMCW transceiver 704 relative to the reference clock 720 of the first FMCW transceiver 702. In some instances, the reference clock frequency drift alpha is caused by pressure, voltage, or temperature variations.

[0081] In some instances, Ethernet-PTP can provide timestamps accurate to within 100 ns or less. In some instances, the maximum acceptable latency of the transmission of the timestamp is responsive to the minimum acceptable accuracy of the range-Doppler information for angle-of-arrival determination. Two Ethernet-PTP timestamps separated by T seconds are provided from the first FMCW transceiver 702 to the second FMCW transceiver 704. The second FMCW transceiver 704 measures the number of cycles of its reference clock that elapse between receipt of the first timestamp and receipt of the second timestamp. This enables the second FMCW transceiver 704 to determine the frequency drift α of the reference clock 722 of the second FMCC transceiver 704 relative to the reference clock 720 of the first FMCW transceiver 702 within (2 × Acc) / T, where Acc is the accuracy of the timestamp. For example, if T equals 4 seconds and the timestamps received by the second FMCW transceiver 704 are each accurate to within 100 ns (Acc = 100 ns), then the drift α of the reference clock 722 of the second FMCC transceiver 704 relative to the reference clock 720 of the first FMCW transceiver 702 can be determined with an accuracy of (2 × 100 ns) / 4 = 0.05 parts per million (ppm). In some instances, the reference clock frequency drift α is determined as shown in Equation 1:

[0082]

[0083] In Equation 1, N a is the number of clock cycles of the reference clock 720 of the first FMCW transceiver 702 that elapse between the first timestamp and the second timestamp. N b is the number of clock cycles of the reference clock 722 of the second FMCW transceiver 704 that elapse between the first timestamp and the second timestamp.

[0084] In some instances, the second FMCW transceiver 704 is modified (e.g., programmed) to compensate for this frequency drift α, as further described below: The start frequency (f 0B ) is modified to be equal to the start frequency (f 0A ) of the first FMCW transceiver 702. (The start frequency refers to the base frequency F0 106 of the signal to be transmitted by each respective transmitter.) The ADC sampling frequency is modified. The start time (ADC start time) at which the ADC circuit 318 samples the IF signal after the corresponding FMCW chirp 102 of the second FMCW transceiver 704 is compensated or corrected. After the initial second FMCW chirp 9081 of the data frame 116, T IC-B (inter-chirp time) is modified for each second FMCW chirp 908 in the data frame 116 to match the T IC-AAlignment (as close as possible in some instances). The inter-frame period 120 of the second FMCW transceiver 704 is adjusted to be aligned with the inter-frame period 120 of the first FMCW transceiver 702.

[0085] The phase of the signal at the virtual antenna is sensitive to the start frequency (F0 106) difference between the transmitting devices. (After FFT processing, the signal at the virtual antenna is represented as a complex number within the corresponding frequency bin. The phase of the signal refers to the phase of the complex number.) Thus, in some instances, relative to f 0A correct f 0B . Assume that α (read as alpha) represents the drift of the reference clock of transceiver 704 relative to the reference clock of transceiver 702, as determined using (e.g.) an Ethernet-PTP signal. Thus, the second FMCW transceiver 704 incorporates an additional frequency offset of αf 0B by programming f 0A as shown in Equation 2:

[0086] f 0B_programmed = f 0A + αf 0A Equation 2

[0087] In Equation 2, f 0A is the desired start frequency, and f 0B_programmed is the programmed start frequency of a particular transmitter 206. As described above, the second FMCW transceiver 704 has a reference clock drift α relative to the first FMCW transceiver 702. This results in a start frequency drift of -αf 0A . Thus, the programmed value of f 0B_programmed enables the actual start frequency of the transmitted FMCW chirp 908 of the second FMCW transceiver 704 to match the start frequency f 0A of the transmitted FMCW chirp 906 of the first FMCW transceiver 702 because the frequency drift and the programmed offset cancel each other out.

[0088] In some instances, the estimation and correction of the reference frequency drift α according to Equation 2 may leave a residual uncorrected reference clock drift α res (measured in ppm). This residual reference clock drift results in a residual phase error θ given by Equation 3 err-res :

[0089]

[0090] In Equation 3, f 0A is the start frequency (expressed in Hz) of the FMCW chirp 906 transmitted by the first FMCW transmitter 702, d is the distance to the object 322 in the distance, and c is the speed of light. In some instances, f0A equals 77 GHz and d equals 100 meters. Thus, knowing α within 0.1 ppm (i.e., α res = 0.1 ppm) achieves a residual phase error θ err-res less than 2°. Thus, accurate knowledge of the reference clock drift α enables the start frequency of the second FMCW transceiver 704 to be corrected to reduce the residual phase error. In some instances, the programming granularity available in the FMCW transceiver to be adjusted (i.e., the first FMCW transceiver 702 or the second FMCW transceiver 704) provides fine tuning of the start frequency f 0B_programmed to achieve a start frequency accuracy of about 10 -3 ppm or better.

[0091] The sampling rate of the ADC circuit 318 of the second FMCW transceiver 704 can be modified to correct for the drift α of its reference clock 722 relative to the reference clock 720 of the first FMCW transceiver 702. In some instances, the ADC sampling rate can be modified as follows to correct for α. For an object 322 in the range at a distance d, the digitized IF frequency ω IF without reference clock drift (α = 0) is given by Equation 4, where S is the slope of the FMCW chirp 102 frequency (frequency change divided by the ramp time) and T S is the ADC sampling period:

[0092]

[0093] A ppm reference clock drift α results in a corresponding digitized IF frequency (1 + α)×ω IF , so the actual digitized IF frequency deviates from the ideal digitized IF frequency by αω IF . This IF frequency deviation αω IF can cause gain and phase mismatches between the corresponding range bins of the virtual antennas across different transceivers. For example, the IF frequency deviation αω IF can cause gain and phase mismatches between the corresponding cells of the amplitude spikes 422 of the Tx1 - Rx1 and Tx3 - Rx3 range - Doppler FFTs 420 (see Figure 5 ), where (for this example) Tx1 - Rx1 is the virtual antenna of the first FMCW transceiver 702 and Tx3 - Rx3 is the virtual antenna of the second FMCW transceiver 704. The phase mismatch θ associated with the IF frequency deviation is determined as shown in Equation 5 err-mm :

[0094]

[0095] In Equation 5, N ADCis the number of samples measured by the ADC circuit 318 based on the FMCW chirp 102, which is equal to the FMCW chirp duration (ramp time 104) divided by T S . In the instance where N ADC is equal to 256 and α is equal to 200 ppm, θ err-mm is approximately 10°.

[0096] The phase mismatch θ can be corrected by modifying (e.g., programming) the sampling period T of the second FMCW transceiver 704 S , as shown in Equation 6, where T err-mm is the modified sampling period: S-MOD T

[0097] T S-MOD = (1 - α) × T S Equation 6

[0098] In some instances, the sampling rate programming granularity (Δ IF = 1 Ksps) of one thousand samples per second is sufficient to reduce the phase mismatch error θ err-min to 1°.

[0099] The ADC start time refers to the time delay between the start of transmission of the line chirp and the first subsequent ADC sampling moment. In some instances, the ADC start time of the second FMCW transceiver 704 can be corrected relative to the ADC start time of the first FMCW transceiver 702 as follows. The ADC start time T start contributes a phase of 2πf IF T start in the virtual antenna signal, where f IF is the IF frequency. Here, the virtual antenna signal refers to the signal value in the range-Doppler FFT 420 corresponding to the virtual antenna; for example, Figure 5 Tx1 - Rx1 in

[0100] Making this phase contribution (2πf IF T start ) the same (or as close as possible) for the virtual antennas of the first FMCW transceiver 702 and the second FMCW transceiver 704 is helpful for accurate angle estimation. That is, the difference in the ADC start time results in a phase difference, which may cause an angle estimation error. In some instances, the drift α between the reference clocks 720 and 722 of the first FMCW transceiver 702 and the second FMCW transceiver 704 corresponds to the phase mismatch 2πf IF αT start related to ADC sampling. In some instances, this phase mismatch can be compensated in various ways.

[0101] First, the phase mismatch depends on the IF frequency; the IF frequency corresponds to the range bin 414 of the range-FFT 412. The phase mismatch can be corrected by performing a phase rotation on the signal values in the corresponding range bin 414 of the range-FFT 412 on the virtual antenna 710 of the second FMCW transceiver 704. (In some instances, on all virtual antennas 710 of the second FMCW transceiver 704.) This phase flip is achieved by multiplying the signal values in the range bin 414 by e -j2πfIFαTstart where j is the square root of -1. In some instances, if α is known to be within 0.1 ppm, using the phase rotation per range bin 414 results in reducing the phase mismatch associated with the ADC sampling to zero or near zero.

[0102] An alternative second method for compensating the phase mismatch associated with the ADC start time involves modifying (e.g., reprogramming) the ADC start time of the second FMCW transceiver 704. In some instances, assuming a maximum IF frequency of 20 MHz, an ADC start time resolution of 0.55 ns is sufficient to achieve a residual IF frequency-related phase mismatch of approximately 2° or less after phase correction.

[0103] In some instances, it is possible to correct T IC-A with respect to T IC-B . The expected (e.g., programmed) start times of the first FMCW chirp 906 and the second FMCW chirp 908 are n × T IC-A (expected to be the same), where n is an integer iterating from zero to (N chirps -1), and where N chirps is the number of chirps in the frame of FMCW chirps to be transmitted by the first FMCW transceiver 702. However, the reference clock frequency drift α causes an inter-chirp timing difference in the second FMCW transceiver 704, such that before correction, T IC-B is equal to n × (1 - α) × T IC-A . The inter-chirp timing difference n × α × T IC-A can result in a phase mismatch between the virtual antennas of the first transceiver 902 and the virtual antennas of the second FMCW transceiver 904 after the Doppler FFT (step 418).

[0104] In a first example method of correcting T IC-B , the inter-chirp time can be modified (e.g., programmed). The inter-chirp time T IC can be modified with a granularity of Δ(T IC-B ), such that if the device is programmed with an integer value L, it produces an inter-chirp time of LΔ(T IC ). (In some instances, Δ(T IC) is equal to 10 ns.) The inter-chirp time on the auxiliary device can be programmed as follows. First, determine a sequence of integers M(n) as shown in Equation 7:

[0105]

[0106] Next, determine a set of subsequent integers L(n) as shown in Equation 8:

[0107] L(n) = M(n) - M(n - 1), n = 1, 2, 3, … N chirps -1 Equation 8

[0108] In Equation 8, L(n) represents the integer value to be programmed into the second FMCW transceiver 704 to control the inter-chirp time T between the nth chirp and the (n - 1)th chirp of the corresponding data frame 116 IC-B . Thus, the inter-chirp time T represented by the digital control using L(n) IC-B will be Δ(T IC ) x L(n).

[0109] In some instances, there will be a residual error in the inter-chirp time, such as a residual error that varies between ±5 ns. This residual error can be regarded as phase noise, which will generate background noise (minimum noise amplitude) after the Doppler FFT. In some instances, this background noise will be approximately 70 dB (dBc) relative to the carrier.

[0110] In a second example method of correcting T IC-B , the residual error remaining after applying the first example method can be corrected by adjusting f 0B . It should be noted that the start frequency change Δf c is equal to the time adjustment of Δf c / S B . In an instance, the start frequency can be programmed with a granularity of 100 Hz or less, where the slope is S B = 10 MHz / μs. In an instance, the granularity of the equivalent time adjustment of T IC-B will be 100 / 10e12 = 10 picoseconds (ps).

[0111] In some instances, compensation for the reference clock frequency drift α as described herein is performed before step 804 of process 800 Figure 8 .

[0112] Modifications may be made to the described embodiments, and other embodiments are possible within the scope of the claims.

[0113] In some instances, processing circuitry other than a DSP is used, such as a central processing unit (CPU).

[0114] In some instances, a spectral estimation technique other than FFT is used to perform range, range-Doppler, and angle spectral estimations, such as a Bartlett beamformer or a minimum variance distortionless response (MVDR) beamformer.

[0115] In some instances, providing range-Doppler spectral information is sufficient to convey corresponding virtual antenna information.

[0116] In some instances, virtual antenna array signals are provided from both the first FMCW transceiver 702 and the second FMCW transceiver 704 to a processor (e.g., the processor at the shared reference clock 712) for processing to determine angle information.

[0117] In some instances, virtual antenna array signals are provided from the first FMCW transceiver 702 to the second FMCW transceiver 704.

[0118] In some instances, a first reference clock 720 is generated independently of the second FMCW transceiver 704. In some instances, a second reference clock is generated independently of the first FMCW transceiver 702. In some instances, a shared reference clock 712 is generated independently of the first FMCW transceiver 702 and the second FMCW transceiver 704.

[0119] In some instances, a frequency-modulated continuous wave (FMCW) radar includes: a reference clock configured to generate a reference clock signal; an FMCW signal generator configured to generate an FMCW chirp; an analog-to-digital converter (ADC) configured to receive the FMCW signal and sample the FMCW signal in response to the reference clock signal to generate FMCW signal samples; and a processor configured to: determine a frequency drift in response to the reference clock signal, a start time, and an end time; determine at least one of the following in response to the frequency drift: an ADC start time, a start frequency of a corresponding FMCW chirp in the FMCW chirps, a chirp slope of a corresponding FMCW chirp in the FMCW chirps, or a corresponding inter-chirp time between corresponding FMCW chirps in the FMCW chirps; receive the FMCW signal samples; and determine a set of virtual antenna signals in response to the FMCW signal samples.

[0120] In this specification, the term "and / or" (when used in the form of, for example, A, B, and / or C) refers to any combination or subset of A, B, and C, such as: (a) only A; (b) only B; (c) only C; (d) A and B; (e) A and C; (f) B and C; and (g) A, B, and C. Additionally, as used herein, the phrase "at least one of A or B" (or "at least one of A and B") refers to an embodiment that includes any one of the following: (a) at least one A; (b) at least one B; and (c) at least one A and at least one B.

[0121] A device "configured to" perform a task or function can be configured by the manufacturer at the time of manufacture (e.g., programmed and / or hardwired) to perform the function and / or can be configured (or reconfigured) by the user after manufacture to perform the function and / or other additional or alternative functions. The configuration can be performed through firmware and / or software programming of the device, through the construction and / or layout of hardware components, and the interconnection of the device, or a combination thereof.

[0122] A circuit or device described herein as including certain components can actually be adapted to be coupled to those components to form the described circuit system or device. For example, a structure described as including multiple functional blocks can alternatively include only functional blocks within a single physical device (e.g., a semiconductor die and / or an integrated circuit (IC) package) and can be adapted to be coupled to at least some of the functional blocks, for example, by an end user and / or a third party, at the time of manufacture or after manufacture to form the described structure.

[0123] The circuits described herein can be reconfigured to include replacement components to provide at least partially similar functionality to that available prior to the component replacement.

[0124] The term "coupled" is used throughout the specification. The term can cover connections, communications, or signal paths that achieve a functional relationship consistent with this specification. For example, if device A provides a signal to control device B to perform an action, then in a first instance, device A is coupled to device B, or in a second instance, device A is coupled to device B through an intermediate component C, provided that the intermediate component C does not substantially change the functional relationship between device A and device B such that device B is controlled by device A via the control signal provided by device A.

[0125] Although some of the elements of the described examples may be included in an IC and other elements are external to the IC, in other examples, additional or fewer features may be incorporated into the IC. Additionally, some or all of the features shown as external to the IC may be included within the IC, and / or some of the features shown as internal to the IC may be incorporated external to the IC. As used herein, the term "IC" means one or more circuits that: (i) are incorporated in / above a semiconductor substrate; (ii) are incorporated in a single semiconductor package; (iii) are incorporated into the same module; and / or (iv) are incorporated in / on the same PCB.

[0126] Unless otherwise stated, "about", "substantially" or "essentially" before a value means + / - 10% of that value, or, if the value is zero, a reasonable range of values near zero.

Claims

1. A frequency modulated continuous wave (FMCW) radar system, comprising: A first FMCW device including a processor configured to receive a first set of FMCW signals corresponding to a field of view (FOV) and process the first set of FMCW signals to generate a first set of range-Doppler spectral information; And A second FMCW device including a processor configured to receive a second set of FMCW signals corresponding to the FOV and process the second set of FMCW signals to generate a second set of range-Doppler spectral information; Wherein the second FMCW device is configured to transmit the second set of range-Doppler spectral information to the first FMCW device; And Wherein the processor of the first FMCW device is configured to determine angle-of-arrival information with respect to one or more objects in the FOV in response to the first and second sets of range-Doppler spectral information.

2. The FMCW radar system according to claim 1, wherein one or more of a fast Fourier transform (FFT), a Bartlett beamformer, or a minimum variance distortionless response (MVDR) beamformer are used to generate the first and second sets of range-Doppler spectral information.

3. The FMCW radar system according to claim 1, wherein the processor of the second FMCW device is configured to perform Doppler compensation with respect to a deliberately included difference in the transmission time between an FMCW chirp corresponding to the first set of range-Doppler spectral information and an FMCW chirp corresponding to the second set of range-Doppler spectral information.

4. A frequency modulated continuous wave (FMCW) radar system, comprising: A first FMCW device, the first FMCW device including: A first FMCW synthesizer configured to generate a first FMCW chirp; A plurality of transmitters configured to transmit the first FMCW chirp into a field of view (FOV); A plurality of receivers configured to receive signals from the FOV to generate a first received signal; And A processor configured to independently process the first received signal from a second FMCW device to generate a first set of virtual antenna array signals; And The second FMCW device configured to be coupled to the first FMCW device, the second FMCW device including: A second FMCW synthesizer configured to generate a second FMCW chirp; A plurality of transmitters configured to transmit the second FMCW chirp into the FOV; A plurality of receivers configured to receive signals from the FOV to generate a second received signal; And A processor configured to independently process the second received signal from the first FMCW device to generate a second set of virtual antenna array signals; Wherein the second FMCW device is configured to transmit the second set of virtual antenna array signals to the first FMCW device; And Wherein the first FMCW device is configured to use the first and second sets of virtual antenna array signals to determine angle-of-arrival information with respect to one or more objects in the FOV.

5. The FMCW radar system according to claim 4, further comprising a shared reference clock circuit configured to be coupled to the first FMCW device and the second FMCW device, the shared reference clock circuit configured to generate a reference clock signal and provide the reference clock signal to the first and second FMCW devices; wherein the first FMCW device and the second FMCW device are both configured to use the reference clock signal to determine one or more of the following: the start frequency of a respective FMCW chirp, the chirp slope of a respective FMCW chirp, the start time of a respective analog-to-digital converter (ADC), or the inter-chirp time.

6. The FMCW radar system according to claim 4, wherein the first FMCW synthesizer includes a reference signal generator configured to generate a first reference clock signal, the first FMCW synthesizer configured to generate the first FMCW chirp in response to the first reference clock signal; and wherein the second FMCW device includes a reference signal generator configured to generate a second reference clock signal, the second FMCW synthesizer configured to generate the second FMCW chirp in response to the second reference clock signal.

7. The FMCW radar system according to claim 6, wherein the first FMCW device is configured to transmit a start time and an end time to the second FMCW device; and wherein the second FMCW device is configured to determine one or more of the following in response to a plurality of clock cycles between the start time and the end time: the start frequency of a respective FMCW chirp, the chirp slope of a respective FMCW chirp, the start time of a respective analog-to-digital converter (ADC), or the inter-chirp time.

8. The FMCW radar system according to claim 7, wherein the start time and the end time correspond to Ethernet-PTP timestamps.

9. The FMCW radar system according to claim 4, wherein the first FMCW device is configured to provide a synchronization pulse to the second FMCW device; and wherein the second FMCW device is configured to determine a data frame start time in response to the synchronization pulse.

10. A method for detecting an object, the method comprising: using a first frequency-modulated continuous wave (FMCW) device and generating a first set of FMCW chirps in response to a first reference clock generated independently of a second FMCW device; using the first FMCW device and transmitting the first FMCW chirps into a field of view (FOV) in response to the first reference clock; using the first FMCW device to receive signals from the FOV to generate a first received signal; and processing the first received signal in response to the first reference clock to generate a first set of virtual antenna array signals; using the second FMCW device and generating a second set of FMCW chirps in response to a second reference clock generated independently of the first FMCW radar device; Transmit the second FMCW chirp into the field of view (FOV) using the second FMCW device and in response to the second reference clock; Receive a signal from the FOV using the second FMCW device to generate a second received signal; And Process the second received signal in response to the second reference clock to generate a second set of virtual antenna array signals; Transmit the second set of virtual antenna array signals from the second FMCW device to the first FMCW device; And Determine angle-of-arrival information relative to one or more objects in the FOV using the first FMCW device in response to the first and second sets of virtual antenna array signals.

11. The method according to claim 10, wherein the first set of virtual antenna array signals includes a first range-Doppler spectrum estimate; and wherein the second set of virtual antenna array signals includes a second range-Doppler spectrum estimate.

12. The method according to claim 10, further comprising performing Doppler compensation using the second FMCW device relative to a deliberately included difference in the transmission time between the FMCW chirp corresponding to the first set of virtual antenna array signals and the FMCW chirp corresponding to the second set of virtual antenna array signals.

13. The method according to claim 10, further comprising: Providing a shared reference clock signal to the first FMCW device and the second FMCW device; wherein both the first FMCW device and the second FMCW device are configured to use the reference clock signal to determine one or more of the following: the start frequency of the respective FMCW chirp, the chirp slope of the respective FMCW chirp, the start time of the respective analog-to-digital converter (ADC), or the inter-chirp time.

14. The method according to claim 10, further comprising: Generating a first reference clock signal using the first FMCW device, wherein the first FMCW device generates the first FMCW chirp in response to the first reference clock signal; And Generating a second reference second reference clock signal using the second FMCW device, wherein the second FMCW device generates the second FMCW chirp in response to the second reference clock signal.

15. The method according to claim 14, further comprising providing a start time and an end time from the first FMCW device to the second FMCW device; wherein the second FMCW device is configured to determine one or more of the following in response to a plurality of clock cycles between the start time and the end time: the start frequency of the respective FMCW chirp, the chirp slope of the respective FMCW chirp, the start time of the respective analog-to-digital converter (ADC), or the inter-chirp time.

16. The method according to claim 15, wherein the start time and the end time correspond to Ethernet-PTP timestamps.

17. The method according to claim 10, further comprising: Providing a synchronization pulse from the first FMCW device to the second FMCW device; And Determine a data frame start time using the second FMCW device in response to the synchronization pulse.

18. A frequency modulated continuous wave (FMCW) radar, comprising: A reference clock configured to generate a reference clock signal; An analog-to-digital converter (ADC) configured to receive an FMCW signal and sample the FMCW signal in response to the reference clock signal to generate FMCW signal samples; And A processor configured to: Provide a start time and an end time to another FMCW radar in response to the reference clock signal; Receive the FMCW signal samples; Determine a first set of virtual antenna signals in response to the FMCW signal samples; Receive a second set of virtual antenna signals from the other FMCW radar; And Determine an angle of arrival in response to the first and second sets of virtual antenna signals.

19. The FMCW radar according to claim 18, wherein the first set of virtual antenna signals includes a first set of range-Doppler spectrum estimates corresponding to the FMCW signal samples, and the second set of virtual antenna signals includes a second set of range-Doppler spectrum estimates not corresponding to the FMCW signal samples.

20. The FMCW radar according to claim 19, wherein one or more of a fast Fourier transform (FFT), a Bartlett beamformer, or a minimum variance distortionless response (MVDR) beamformer is used to generate the first and second sets of range-Doppler spectrum estimates.

Images