Frequency estimation system and method for estimating coherent distance

Abstract

The distance estimation method includes transmitting a radiation wave modulated in the frequency domain by an emitter into a scene, receiving a reflection of the transmitted wave from the scene, and interfering a copy of the transmitted wave with the received reflection to generate a series of samples of a beat signal having a wrapped phase in the time domain. The method also includes iteratively estimating the frequency of the beat signal in the time domain until a termination condition is met. The iterative estimation of the frequency of the beat signal is based on phase unwrapping of the samples of the beat signal, which are subject to correlated phase errors resulting from the emitter's phase noise statistics, and linear regression that fits the frequency of the beat signal to the unwrapped phase of the beat signal.

Description

[Technical Field]

[0001] The present disclosure relates generally to detecting objects in a scene using frequency modulated continuous wave (FMCW) techniques, and more particularly to an FMCW-based apparatus and method for estimating coherent distances of objects in a scene. [Background technology]

[0002] Several application fields, including autonomous vehicles, industrial robotics, navigation, aerospace, meteorological element measurement, and air quality monitoring, require accurate object detection to perform critical functions. Reliable object detection is typically performed using remote sensing techniques such as Light Detection and Ranging (LIDAR), which uses pulsed laser light to measure variable distances to objects in a scene. Currently, several types of LIDAR are available to estimate the distance of objects in a target scene. Traditional pulse LIDARs transmit short pulses of light toward an object and calculate the distance to the object from the elapsed time it takes for the reflected light to return to the LIDAR system. Pulse LIDARs are incoherent and only measure the intensity of light reaching the detector, making them susceptible to errors due to ambient light or interference from other LIDARs.

[0003] An alternative method for estimating distance using laser light is frequency-modulated continuous wave (FMCW) lidar. Unlike pulsed lidar, FMCW lidar has a constant illumination power, making it compatible with integrated photonics. FMCW lidar is a coherent ranging technique that measures distance by mixing a local copy of the transmitted laser beam with the reflected light returning to the receiver. The beat frequency of the resulting interference signal is proportional to the distance of the reflecting object, making distance measurement a frequency estimation problem.

[0004] Despite these advantages, the accuracy of FMCW lidar is often limited by the phase noise of the laser source. Lasers cannot generate light at a single pure frequency through fundamental oscillations. Instead, thermal changes, mechanical vibrations, and even the quantum nature of photons introduce randomness into the oscillation phase, corresponding to deviations in the emission frequency. In FMCW lidar applications using noisy lasers, the laser's phase noise similarly causes the interference signal to deviate from the true beat frequency. Because these deviations increase with object distance, traditionally there is a maximum distance (coherence distance) that can be measured without excessive error. This understanding is based on classical frequency estimation techniques, which assume a sinusoid in additive noise, and the maximum likelihood estimate of the frequency of the sinusoid in additive noise is the peak of the periodogram. However, methods for determining these peaks do not work well when phase noise distributes the signal power across a range of frequencies.

[0005] Conventional techniques estimate the frequency in the frequency domain. However, this approach in conventional methods suffers from shortcomings because phase noise statistics are not well understood in the frequency domain. Attempts to decipher the phase noise statistics increase the complexity of the measurement system and / or require more hardware and computational resources. Therefore, a better approach is needed to estimate the frequency of a beat signal. Summary of the Invention

[0006] It is an object of some embodiments to provide an improved frequency estimation technique for determining the distance of an object in a target scene. It is also an object of some embodiments to provide a system and method for performing frequency estimation in the time domain, which explicitly models phase noise statistics so that they are better described and understood. Some exemplary embodiments aim to perform frequency estimation in the time domain and take advantage of known phase noise statistics. Some exemplary embodiments are particularly directed to providing a frequency estimation technique for scenarios where the maximum measurable distance is greater than the coherence length of the light source.

[0007] Some exemplary embodiments recognize that in coherent ranging techniques, the distance to an object in a scene is a function of the beat frequency of an interference signal resulting from mixing a local copy of the transmitted beam with the reflected light returning to the receiver, thereby reducing the problem of measuring object distance to a frequency estimation problem. In coherent ranging techniques, such as FMCW, depth measurements are affected by correlated noise. Therefore, some exemplary embodiments recognize that in the case of FMCW lidar, laser phase noise causes the interference signal to deviate from the true beat frequency, making beat frequency estimation prone to errors if phase noise statistics are not considered. In this regard, some exemplary embodiments recognize that phase noise statistics, which are not well understood in the frequency domain, can be explicitly modeled in the time domain.

[0008] Some exemplary embodiments recognize that frequency estimation in the time domain can be performed by a two-stage process involving phase unwrapping and linear regression. In this regard, some exemplary embodiments consider prior information regarding the linear nature of the underlying unwrapped phase. Some exemplary embodiments recognize that FMCW depth measurements are subject to correlated noise and therefore incorporate phase error estimation into the alternating optimization. To this end, the exemplary embodiments estimate frequency in terms of phase unwrapping, explicitly taking into account the correlation of the phase noise.

[0009] Additionally, some exemplary embodiments aim to determine the most likely unwrapping sequence of the wrapped phase using a linear minimum mean square error estimate of the phase error. Some exemplary embodiments recognize that phase error statistics are easier to derive than the power spectral density (PSD) resulting from a noisy sinusoidal measurement. Thus, exemplary embodiments of the present disclosure can apply non-white laser frequency noise distributions to frequency estimation.

[0010] Various exemplary embodiments consider a laser light source for frequency estimation as part of a larger distance estimation problem. In this area, FMCW lidar is more robust than pulsed lidar to interference from ambient light or other lidar systems, making it promising for applications such as autonomous navigation. However, as discussed above, phase noise in FMCW lidar causes the instantaneous frequency to deviate from the desired frequency modulation, reducing the temporal coherence of the interfering light beam. Some exemplary embodiments recognize that the loss of temporal coherence becomes more severe as the distance to the target increases. The coherence distance, determined by the amount of phase noise, is considered to be the distance beyond which ranging cannot be reliably performed. Some exemplary embodiments also recognize that to reach long distances (e.g., those required for navigation), a hardware solution is to use lasers with small linewidths (i.e., small phase noise). However, such small linewidth lasers tend to be expensive and increase the overall system hardware complexity.

[0011] Some exemplary embodiments recognize that one attempt to increase range can be made by considering the effects of phase noise. For example, some exemplary embodiments recognize that, assuming a white frequency noise laser model, a Lorentzian distribution can be fitted to the power spectral density (PSD) of the measured interference signal because the PSD is asymptotically Lorentzian as a function of range. Some other exemplary embodiments also recognize that to address the problem of phase noise affecting lidar measurements, one solution is to perform a frequency-domain estimation by fitting a Lorentzian curve to the power spectral density of the measurement. However, because the Lorentzian curve is an approximation of the interference PSD, fitting assuming Gaussian statistics is not optimal.

[0012] Additionally, some other illustrative embodiments recognize that to address the issue of phase noise affecting lidar measurements, another solution is to measure the characteristics of phase impairments (including phase noise and nonlinearities) using an additional reference arm and fixed depth. In particular, compensation for impairments can be achieved instead of including impairment modeling in the estimation process (thus using standard frequency-domain peak measurements). However, such a solution requires additional hardware and calibration.

[0013] In light of the above realizations, various exemplary embodiments are based on the understanding that attempts to correct for phase errors or to fit the PSD with a curve that makes simple assumptions about the distribution will result in suboptimal or infeasible solutions. Furthermore, some exemplary embodiments recognize that such approaches cannot achieve an acceptable level of accuracy, especially under high signal-to-noise ratio conditions. Toward these ends, some exemplary embodiments propose using accurate phase noise statistics in the time domain to achieve better performance, especially at high SNRs. Furthermore, the time-domain method provided by some exemplary embodiments uses statistics from a realistic laser model instead of a naive white-frequency noise model.

[0014] Therefore, one key issue for some exemplary embodiments is how to best account for the impact of phase noise in the ranging process. In this regard, some exemplary embodiments recognize that it is easiest to directly describe phase noise statistics in the time domain (i.e., in terms of the phase of the interfering signal) rather than in the frequency domain after a Fourier transform. Specifically, some exemplary embodiments describe the interfering signal phase error in a closed form for both white frequency noise and colored frequency noise, compared to the PSD, which is only known in a closed form for white frequency noise. Some exemplary embodiments recognize that the phase error is a stationary Gaussian process and therefore is completely described by the autocorrelation function. Some exemplary embodiments are based on the understanding that a quadrature demodulator can be used to extract the interfering signal phase to perform phase-based frequency estimation. While FMCW lidar systems traditionally use one balanced detector, a quadrature demodulator includes two balanced detectors to directly perform in-phase and quadrature measurements.

[0015] Some exemplary embodiments provide a solution for frequency estimation via phase unwrapping. Some exemplary embodiments utilize phase error correlation to unwrap the measured phase and then perform linear regression of the unwrapped phase, again using the phase error correlation. In this regard, some exemplary embodiments provide an iterative algorithm based on the principles described above. This iterative algorithm is initialized and refined by less rigorous methods (e.g., Lorentzian fitting) and exhibits better accuracy at high SNRs.

[0016] To achieve the foregoing objects and advantages, some illustrative embodiments provide a system, method, and program for estimating coherent distances of objects in a scene.

[0017] For example, some exemplary embodiments provide an FMCW device. The FMCW device includes an emitter configured to transmit at least one radiation wave into a scene, the transmission wave being modulated in the frequency domain using a linear modulation that is impaired to cause nonlinearity of the transmission wave in the frequency domain. A receiver receives a reflection of the transmission wave from the scene, and a mixer interferes a copy of the transmission wave with the received reflection of the transmission wave to generate a beat signal. A pair of analog-to-digital converters generates a series of samples of the beat signal having a wrapped phase in the time domain. A processor is configured to iteratively estimate the frequency of the beat signal until a termination condition is met. The iterative estimation of the frequency of the beat signal is based on phase unwrapping of the samples of the beat signal that are subject to correlated phase errors derived from the emitter's phase noise statistics and linear regression that fits the frequency of the beat signal to the unwrapped phase of the beat signal. The estimated frequency of the beat signal may be utilized by a circuit to estimate a distance to an object in the scene.

[0018] Some exemplary embodiments also provide a frequency estimation method utilizing an FMCW device. The method includes transmitting at least one radiation wave into a scene, the transmitted wave being linearly modulated in the frequency domain, which is subject to impairments that cause nonlinearity of the transmitted wave in the frequency domain. The method further includes receiving a reflection of the transmitted wave from the scene, interfering a copy of the transmitted wave with the received reflection of the transmitted wave to generate a beat signal, and generating a series of samples of the beat signal to have a wrapped phase in the time domain. The method also includes iteratively estimating the frequency of the beat signal in the time domain until a termination condition is met. The iterative estimation of the frequency of the beat signal is based on phase unwrapping of the samples of the beat signal subject to correlated phase errors derived from emitter phase noise statistics and linear regression that fits the frequency of the beat signal to the unwrapped phase of the beat signal.

[0019] Some exemplary embodiments also provide a non-transitory computer-readable medium having stored thereon instructions executable by a computer for performing a frequency estimation method using an FMCW device. The frequency estimation method includes transmitting at least one radiation wave into a scene, the transmitted wave being linearly modulated in the frequency domain, which is subject to impairments that cause nonlinearity of the transmitted wave in the frequency domain. The method further includes receiving a reflection of the transmitted wave from the scene, interfering a copy of the transmitted wave with the received reflection of the transmitted wave to generate a beat signal, and generating a series of samples of the beat signal to have a wrapped phase in the time domain. The method also includes iteratively estimating the frequency of the beat signal in the time domain until a termination condition is met. The iterative estimation of the frequency of the beat signal is based on phase unwrapping of the samples of the beat signal, which are subject to correlated phase errors derived from emitter phase noise statistics, and linear regression that fits the frequency of the beat signal to the unwrapped phase of the beat signal.

[0020] According to some exemplary embodiments, the generated beat signal is distorted by nonlinearities in the linear modulation caused by the impairment. In some exemplary embodiments, the current iteration of the iterative estimation includes, for each sample of the series of samples of the beat signal, determining a current phase error and a phase unwrapping number that fit to a previous frequency of the beat signal and a previous phase offset of the beat signal determined during a previous iteration. The current phase error of the current sample in the series of samples of the beat signal is correlated with a previous phase error of a previous sample in the series of samples of the beat signal by a predetermined phase noise statistic. The current iteration of the iterative estimation also includes updating the current frequency of the beat signal and the current phase offset of the beat signal during the current iteration based on the determined current phase error and the determined phase unwrapping number.

[0021] According to some exemplary embodiments, the iterative frequency estimation is based on an alternating optimization including a Viterbi algorithm that probabilistically determines a current phase error and a phase unwrapping number to maximize fitting likelihood across a series of samples of the beat signal. The Viterbi algorithm uses the determined previous phase error and previous phase unwrapping number of previous samples to perform causal estimation of the current phase error and the current phase unwrapping number of each current sample of the series of samples of the beat signal. The causal estimation of the current phase error and the current phase unwrapping number is determined via linear minimum mean square error estimation.

[0022] Embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings, which are not necessarily to scale, emphasis instead being placed upon illustrating the principles of embodiments of the present disclosure. [Brief explanation of the drawings]

[0023] [Figure 1A] FIG. 1 illustrates a workflow for estimating depth of a scene using the principles of a frequency modulated continuous wave device, according to some exemplary embodiments. [Figure 1B] FIG. 1 is a schematic diagram illustrating an FMCW lidar system for generating estimates of distances to objects, in accordance with some illustrative embodiments. [Figure 2A] FIG. 1 is a detailed schematic diagram illustrating an FMCW lidar system, according to some embodiments. [Figure 2B] FIG. 1 illustrates a swept-frequency signal emitted from a tunable laser of an FMCW lidar system and a delayed copy captured by a receiver, in accordance with some illustrative embodiments. [Figure 2C] FIG. 1 illustrates an example method for estimating distance using an FMCW lidar system, in accordance with some example embodiments. [Figure 3] FIG. 1 illustrates an iterative frequency estimation method utilizing principles of an FMCW lidar system, in accordance with some illustrative embodiments. [Figure 4] 4 illustrates a method for illustrating one iteration of the iterative frequency estimation method of FIG. 3 in accordance with some exemplary embodiments. [Figure 5] FIG. 1 illustrates a Viterbi phase unwrapping algorithm for estimating frequency in accordance with some exemplary embodiments. [Figure 6] FIG. 1 is a schematic diagram illustrating a process for estimating frequency based on signal phase in accordance with some exemplary embodiments. [Figure 7A] FIG. 2 is a detailed schematic diagram illustrating the phase unwrapping step of the frequency estimation algorithm in accordance with some exemplary embodiments. [Figure 7B] FIG. 2 is a detailed schematic diagram illustrating the phase unwrapping step of the frequency estimation algorithm in accordance with some exemplary embodiments. [Figure 7C] FIG. 1 illustrates a process for refining frequency via linear regression in a frequency estimation algorithm in accordance with some demonstrative embodiments. [Figure 8] 1 is a scenario illustrating an example use case of an FMCW lidar system, according to some example embodiments. [Figure 9] FIG. 1 is a block diagram illustrating a system for implementing an FMCW lidar system, in accordance with some illustrative embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0024] While the above-described drawings illustrate embodiments of the present disclosure, other embodiments are contemplated, as discussed above. The present disclosure illustrates exemplary embodiments and is not intended to be limiting. Those skilled in the art can devise numerous other variations and implementations that fall within the scope and spirit of the principles of the disclosed embodiments.

[0025] The following description provides exemplary embodiments only and is not intended to limit the scope, application, or configuration of the present disclosure. Rather, the following description of exemplary embodiments will provide those skilled in the art with an enabling description for implementing one or more exemplary embodiments. Various changes may be made in the function and arrangement of elements without departing from the spirit and scope of the subject matter as set forth in the appended claims.

[0026] In the following description, specific details are given to provide a thorough understanding of the embodiments. However, those skilled in the art will understand that the embodiments can be practiced without these specific details. For example, systems, processes, and other elements in the disclosed subject matter may be shown as components in block diagrams so as not to obscure the embodiments in unnecessary detail. Also, well-known processes, structures, and techniques may be shown without unnecessary detail so as not to obscure the embodiments. Furthermore, like reference numbers and names in the various drawings refer to like elements.

[0027] Each embodiment may also be described as a process, which is depicted as a flowchart, flow diagram, data flow diagram, structure diagram, or block diagram. Although a flowchart may describe operations as a sequential process, many operations may be performed in parallel or simultaneously. The order of operations may also be changed. A process may be terminated when its operations are completed, but the process may include additional steps not discussed or shown. Furthermore, not all operations within a specifically described process need be included in all embodiments. A process may be a method, a function, a procedure, a subroutine, a subprogram, etc. When a process is a function, the termination of the function corresponds to the function returning to the calling function or the main function.

[0028] Furthermore, embodiments of the disclosed subject matter may be implemented, at least in part, manually or automatically. The manual or automatic implementation may be implemented, or at least assisted, by machine, hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When implemented by software, firmware, middleware, or microcode, the program code or code segments to perform the necessary tasks may be stored on a machine-readable medium. A processor may perform the necessary tasks.

[0029] Imaging techniques are widely adopted in several applications due to the need to understand the environment around an object in order to make informed decisions. For example, in remote sensing applications, it is often desirable to describe the physical characteristics of an area as clearly as possible. Also, in other applications, such as autonomous and semi-autonomous vehicles, making time-bound decisions is important for the operation of these vehicles. In all of these applications, it is essential to accurately perform depth estimation of objects within the target scene. Accurate depth estimation can be achieved by accounting for errors that occur when measuring the scene. In many of these scenarios, illumination sources often introduce errors into measurements.

[0030] Lidar (lidar) is an increasingly popular sensing modality in a variety of ranging applications, from autonomous driving to industrial robotics and lunar navigation. Traditional pulsed lidars transmit short pulses of light toward an object and calculate the distance to the object from the elapsed time it takes for the reflected light to return to the lidar system. Pulsed lidars are incoherent and measure only the intensity of the light reaching the detector, making them susceptible to errors from ambient light or interference from other lidars. An alternative method for estimating distance using laser light is frequency-modulated continuous wave (FMCW) lidar. Unlike pulsed lidar, FMCW lidar has a constant illumination power, making it compatible with integrated photonics. FMCW lidar is a coherent ranging technique that measures distance by mixing a local copy of the transmitted laser beam with the reflected light returning to the receiver. Because the beat frequency of the resulting interference signal is proportional to the distance of the reflecting object, measuring distance becomes a frequency estimation problem.

[0031] However, like any oscillator, a laser cannot generate light at a single pure frequency. Instead, thermal variations, mechanical vibrations, and even the quantum nature of photons introduce randomness into the phase of the oscillation, corresponding to deviations in the emitted frequency. When using a laser for FMCW lidar, the laser's phase noise similarly causes the interference signal to deviate from the true beat frequency. Typically, assuming a sinusoid in additive noise, there is a maximum distance that can be measured without excessive error. The maximum likelihood estimate of the frequency of a sinusoid in additive noise is the peak of the periodogram, but peak-finding methods do not work well when the phase noise distributes the signal power across a range of frequencies.

[0032] Even when the distance to an object is greater than the coherence length of the laser, FMCW lidar measurements still contain distance information. However, modern techniques are needed to consider the phase noise statistics when performing frequency estimation. Generally, phase noise statistics can be explicitly modeled in the time domain. However, frequency estimation is typically performed in the frequency domain, where phase noise statistics are less well understood. Instead, when the maximum distance to be measured is greater than the coherence length of the laser, it is effective to perform frequency estimation in the time domain and take advantage of known phase noise statistics.

[0033] Some solutions aim to perform frequency estimation in the time domain by phase unwrapping and linear regression. Many of these algorithms perform a naive phase unwrapping procedure that does not consider the linearity of the underlying signal. More sophisticated methods use prior knowledge that the underlying unwrapped phase is linear in algorithms that alternate between phase unwrapping using frequency and phase offset estimates and updating frequency and phase offset estimates that take the estimated unwrapping into account. However, none of these methods consider the presence of correlated noise. Because FMCW depth measurements are affected by correlated noise, the exemplary embodiments disclosed herein incorporate phase error estimation into the alternating optimization.

[0034] Given a candidate frequency, the proposed solution uses the Viterbi algorithm and phase noise statistics to approximately recover the maximum likelihood unwrapping sequence. The algorithm then alternates between unwrapping and refining the frequency estimate until convergence is achieved. The proposed solution consistently achieves excellent performance over long distances or with large linewidth lasers when the signal-to-noise ratio is sufficiently high. These and several other advantages result from the systems, methods, and programs disclosed by the exemplary embodiments.

[0035] A frequency-modulated continuous-wave lidar, or FMCW lidar, system is a specialized lidar system that measures both the range and velocity of moving objects. Such measurements are achieved by continuously varying the frequency of a transmitted signal over a fixed period at a known rate using a modulated signal. This modulation is often used to measure distance very accurately at close ranges by comparing the phase of the frequencies of two echo signals. FIG. 1A illustrates a workflow 100A for estimating the depth of a scene using the principles of a frequency-modulated continuous-wave (FMCW) system, according to some exemplary embodiments.

[0036] The transmit waves transmitted into the scene are generated by one or more lasers in the lidar and launched toward the scene by an appropriate transmitter in the FMCW lidar system. A copy of the transmit waves 10 is stored locally for further processing. The transmit waves are reflected from the scene, and a reflected wave 20 is collected by an appropriate receiver in the FMCW lidar system. The transmit wave copies 10 and the reflected wave 20 are fed into a signal-mixing mixer 30, which outputs a measurement of the resulting interference signal. As part of the mixing process, the optical signal is converted to an analog electronic signal, which is then digitized to generate a sample of the interference signal.

[0037]

number

[0038]

number

[0039] In-phase interference measurements 40 and quadrature measurements 50 are obtained as outputs of signal mixing process 30. In some exemplary embodiments, an FMCW receiver can obtain in-phase interference measurements 40 using one balanced detector, and quadrature measurements 50 are calculated from in-phase interference measurements 40 via a Hilbert transform. In some exemplary embodiments, a quadrature demodulator including two pairs of balanced detectors can obtain both in-phase measurements 40 and quadrature measurements 50. In-phase measurements 40 are given by:

number

[0040]

number

[0041]

number

[0042]

number

[0043] The resulting equation expresses the interfering signal in terms of its amplitude, beat frequency, phase offset, a stationary random process representing the phase change, and a circularly symmetric complex AWGN with autocorrelation. The unwrapped phase is then defined as a function of the beat frequency, phase offset, and total phase error. The unwrapped phase is the sum of the extracted wrapped phase and an unknown integer number of cycles that must be added to the wrapped phase to perform phase unwrapping.

[0044] Depth estimation 70 requires estimating the frequency of the beat signal from the wrapped phase vector using known phase change statistics 60. The final depth estimation then rescales the frequency estimate using the chirp rate γ and the speed of light c as follows:

number

[0045]

number

[0046]

number

[0047] After obtaining the frequency of the beat signal, a distance estimate to the object that reflected the reflected wave 20 may be determined as an output of depth estimation 70. In this manner, exemplary workflow 100A provides an accurate estimate of the beat frequency even in the presence of significant phase errors by taking into account correlated noise.

[0048] The exemplary workflow 100A may be compiled as a machine-executable process or method and may be executed using an FMCW device or lidar system. Details of the components and operation of such an FMCW lidar system are described below with reference to FIG. 1B. FIG. 1B is a schematic diagram illustrating an FMCW lidar system 100B that generates an estimate 118 of a distance to an object 106. The FMCW lidar system 100B includes an emitter 102, a transmitter 104, a receiver 108, a mixer 110, an analog-to-digital converter (ADC) 112, a processor 114, and a memory 116.

[0049] In some exemplary embodiments, the emitter 102 includes a suitable light source, such as a laser, and frequency modulation electronics. In some embodiments, the laser may be a tunable laser, and the frequency may be modulated by directly controlling the lasing. In some exemplary embodiments, the laser may be a fixed-frequency laser, and the frequency may be modulated externally, for example, using an electro-optic modulator. In some exemplary embodiments, the emitter 102 transmits radiation (e.g., a laser) into the scene. The radiation is linearly modulated in the frequency domain before being transmitted. The linear modulation is affected by impairments, for example, due to source phase noise, which causes nonlinearities in the radiation in the frequency domain.

[0050] In some exemplary embodiments, the transmitter 104 includes optics to direct and focus the laser beam onto the measurement target object 106. Focusing may be achieved via a lens or a combination of lenses. Beam steering techniques utilized by some exemplary embodiments include, but are not limited to, mechanically scanned mirrors, optical phased arrays, microelectromechanical systems (MEMS) mirrors, and liquid crystal metasurfaces.

[0051] The receiver 108 includes focusing optics for collecting the light reflected from the object, and may include a lens or group of lenses for focusing the reflected light, and a free-space-to-fiber coupler for coupling the received light to a fiber optic cable.

[0052] The mixer 110 generates a beat signal by optically combining a copy of the light from the emitter 102 (local oscillator) with the light captured by the receiver. The beat signal is distorted by the nonlinearity of the linear modulation caused by impairments due to laser phase noise. In some embodiments, the mixing is performed using a beam splitter. In other embodiments, the mixing is performed by an optical fiber coupler. A detector in the mixer 110 then converts the optical signal into an analog electronic signal (e.g., voltage or current). The detector can be a single photodiode or a group of photodiodes, such as a balanced detector. The detector can also include an amplifier circuit to increase the level of the electrical signal.

[0053] The ADC 112 converts the analog electrical signal to a digital signal at predetermined sampling times. The ADC 112 includes sampling and quantization electronics that output the digital signal. In some embodiments, the ADC 112 samples at evenly spaced times. In other embodiments, the ADC 112 uses an external reference, such as a k-clock, to determine the sampling time that corresponds to the linear frequency sweep of the laser.

[0054] The processor 114 uses the digitized samples of the mixed signal to estimate the distance to the object. In some embodiments, the processor 114 is an on-board device such as a field programmable gate array (FPGA). In other embodiments, the processor 114 is an external computer. The memory 116 stores, among other things, calibration data and parameters necessary for the processor to generate accurate distance estimates 118.

[0055] FIG. 2A is a detailed schematic diagram illustrating an FMCW lidar system 200, according to some embodiments. In some exemplary embodiments, the FMCW lidar system 200 may be embodied as an apparatus. In some exemplary embodiments, the FMCW lidar system 200 may be a distributed system including electrical and optical devices. For example, the FMCW lidar system 200 may be implemented using a controller and optoelectronic devices. The controller may include data processing elements of the FMCW lidar system 200, such as a processor and associated circuitry. The optoelectronic elements may include other electrical and optical elements of the FMCW lidar system 200.

[0056] The FMCW lidar system 200 includes an emitter 202, a transmitter 204, a receiver 208, a mixer 210, a pair of analog-to-digital converters (ADCs) 212A and 212B, and a processor 214. The emitter 202 includes a tunable laser 222 and a modulation controller 220. The tunable laser 222 is connected via a fiber optic coupling to a splitter 224, which generates two copies of the laser beam. The first copy is sent to the transmitter 204, where a collimator 226 focuses the light from the optical fiber into a parallel beam. Beam steering optics, such as a pair of galvo mirrors 228, direct the transmitted beam toward the object 206.

[0057] The wavelength of the tunable laser 222 is swept by the modulation controller 220 so that the optical frequency of the laser beam is a linear function of time (i.e., linearly modulated in the frequency domain). As shown in Figure 2B, according to some exemplary embodiments, a swept frequency signal is emitted from the tunable laser 222 and a delayed copy of it is captured by the receiver 208 of the FMCW lidar system 200. The reflected light from the object 206 is captured by the receiver 208, which focuses the light from free space into an optical fiber.

[0058]

number

[0059]

number

[0060] Referring to FIG. 2A, mixer 210 includes a 90° optical hybrid 232 and a pair of balanced detectors 230A and 230B. Mixer 210 receives receive light (RX) through an input port and a local oscillator (LO), a second copy of the laser beam, through another input port. The LO and RX fields are combined in optical hybrid 232 to generate two pairs of mixed outputs. Each output is connected to a balanced detector (230A or 230B), which converts the optical signal to an electrical signal using a pair of photodiodes and amplifiers. In this regard, mixer 210 has one branch for the in-phase component (I) and one branch for the quadrature component (Q). In the I branch, the LO and RX optical signals are combined to generate a photocurrent in each photodiode of balanced detector 230A, as given by the following equation:

number

[0061]

number

[0062]

number

[0063] The LO and RX optical signals are then combined to generate a photocurrent in each photodiode of balanced detector 230B given by the following equation:

number

[0064] The balanced detector 230B filters out the DC term and common mode noise by taking the difference between the two photocurrents.

number

[0065] The orthogonal measurements take the form:

number

[0066]

number

[0067] Each ADC (212A or 212B) includes circuitry for low-pass filtering, sampling, and quantizing an electrical signal to generate in-phase (I) and quadrature (Q) digital measurements. Processor 214 extracts the phase from the I / Q components to provide depth estimation.

[0068] In some exemplary embodiments, the FMCW lidar system and the FMCW device may be the same. In some alternative embodiments, the lidar system may be external to the FMCW device, and the FMCW device may control one or more operations of the lidar. Regardless of the implementation, FMCW lidar system 200 may perform workflow 100A as a process or method. FIG. 2C illustrates one exemplary method 250 for estimating distance using an FMCW lidar system, such as FMCW lidar system 200, according to some exemplary embodiments.

[0069] Method 250 includes step (3) of splitting incident light into a local beam and a transmit beam. Splitting step 3 may be performed by emitter 202 of FMCW lidar system 200. The transmit beam is transmitted (5) toward a target scene by transmitter 204 of FMCW lidar system 200. A reflection of the transmit beam from the scene is received (7) by receiver 208 of the FMCW lidar system as a reflected beam. One or more objects in the scene can reflect the transmit beam toward the FMCW lidar system. Mixer 210 generates an interference pattern (9) by interfering the reflected beam received from the scene with the local beam. Thus, mixer 210 can perform interference according to the principles of signal mixing process 30 of workflow 100A.

[0070] One or more analog-to-digital converters (ADCs) 112 generate interference signal samples by sampling the interference pattern at linearly spaced frequencies (11) in the manner described with reference to workflow 100A of FIG. 1A. One or more processors 114 interface with memory 116 storing executable programs, data, and instructions to determine distances to objects in the scene via joint phase unwrapping and linear regression of the phase extracted from the interference samples (13). Refinement via phase unwrapping and linear regression may be performed in the manner described with reference to workflow 100A. Distance is determined by rescaling the estimated beat frequency with the chirp rate and the speed of light. One or more processors 114 can control an interface 120, such as an output device, to output the determined object distances in any suitable manner (15).

[0071] FIG. 3 illustrates an iterative frequency estimation method 300 utilizing components of an FMCW device, according to some exemplary embodiments. The frequency estimation method 300 may be compiled and executed as a control process of an FMCW lidar system. In some exemplary embodiments, the method 300 may include operations performed by one or more components of the FMCW lidar system 200. The method 300 includes a step (301) of transmitting at least one radiation wave toward a scene. The transmitting step 301 may be preceded by preprocessing, such as generating and directing the at least one radiation wave, which may be performed by any suitable illumination source, such as a laser, a UV lamp, an IR lamp, or the like. The preprocessing may also include obtaining a local copy of the transmitted wave. The transmitting step 301 may include guiding the at least one radiation wave toward the scene using appropriate optics and associated electronics.

[0072] The scene may include one or more objects, which, when illuminated, may return a reflection of the illuminated transmitted wave. One or more receivers receive (303) a reflection of the transmitted wave from the scene. Method 300 further includes generating (305) a beat signal by interfering a copy of the transmitted wave with a received reflection of the transmitted wave. The interference may be performed in the manner previously described with respect to FIGS. 1A and 2A. The interference generates (307) the beat signal as a series of samples with a wrapped phase in the time domain. The frequency of the beat signal in the time domain is then iteratively estimated (309) by iteratively performing phase unwrapping followed by linear regression until a termination condition is met.

[0073]

number

[0074]

number

[0075]

number

[0076] A wrap phase 604 is extracted from the measured signal 602. However, the wrap phase may be insufficient to estimate frequency in the time domain because it lacks information about where and how often wrapping occurs. Predicting the location and magnitude of phase wrapping is further complicated by deviations due to phase noise. Therefore, to perform robust frequency estimation, a frequency estimation algorithm 612 jointly and alternately unwraps the phase and estimates the frequency. In addition to the wrap phase 604, the estimation algorithm 612 requires knowledge of phase noise statistics 606 and an initial frequency estimate 610. The phase noise statistics 606 are known from prior calibration or manufacturer characteristics specifications such as linewidth. The initial frequency estimate 610 is obtained by calculating the power spectral density of the measured signal 602 via a Fourier transform.

[0077] Given these inputs, the frequency estimation algorithm 612 alternates between two steps: phase unwrapping 614, which predicts the location and magnitude of phase wrapping, and linear regression 616, which updates the frequency estimate. After a sufficient number of iterations, the latest result of the linear regression is taken as the final frequency estimate 618.

[0078]

number

[0079]

number

[0080]

number

[0081]

number

[0082]

number

[0083]

number

[0084]

number

[0085] In some embodiments, the phase error estimate is a linear minimum mean square error (LMMSE) estimate given by:

number

[0086]

number

[0087]

number

[0088]

number

[0089]

number

[0090]

number

[0091]

number

[0092]

number

[0093]

number

[0094]

number

[0095]

number

[0096]

number

[0097]

number

[0098] Some other important aspects of the frequency estimation algorithm 612 of FIG. 6 are now described.

[0099] Initialization

number

[0100] Noise Statistics

number

[0101]

number

[0102]

number

[0103]

number

[0104] 8 is a scenario illustrating an example use case of an FMCW lidar system, according to some demonstrative embodiments. A vehicle 802 traveling on a road link 804 may be connected to a base station 806. The road link may have one or more obstacles 808 present, and the goal is to traverse the road link 804 without colliding with the obstacles 808. Therefore, to plan a path for the vehicle 802 to avoid the obstacles 808, an accurate estimation of the depth / distance of the obstacles 808 is required.

[0105] The vehicle 802 may be a manually operated vehicle, a semi-autonomous vehicle, or a fully autonomous vehicle, and may be configured to communicate with a base station 806. Accordingly, the vehicle 802 may include suitable elements for performing remote sensing, data communication, and data processing. To this end, the vehicle may be considered to include an onboard FMCW lidar system, as described above. The lidar system emitter transmits a radiated wave or beam (shown as a solid line) toward the obstacle 808 and receives a reflected wave (shown as a dotted line) of the transmitted wave / beam from the scene ahead. Due to the vehicle 802 moving and / or errors in the light source laser, the depth of the obstacle 808 may not be accurately estimated using conventional techniques.

[0106] The vehicle 802 can perform an accurate depth estimation of the obstacle 808 by calling the FMCW lidar system. The FMCW lidar system may be fully or partially mounted on the vehicle 802. In embodiments where the FMCW lidar system is partially mounted on the vehicle 802, the computational step of performing the frequency estimation may be performed via edge computing at the base station 806. In some exemplary embodiments, the vehicle 802 can determine the geolocation of the obstacle 808 by determining the precise distance of the obstacle 808 via the FMCW lidar system. An onboard controller of the vehicle 808 can then alter the path of the vehicle 808 to avoid a collision with the obstacle 808. In some exemplary embodiments, additionally or alternatively, the vehicle 802 can communicate at least the geolocation of the obstacle 808 to the base station 806 to update a map of the area in which the road link 804 is located.

[0107] Although the exemplary use cases are described with reference to road vehicles, the exemplary embodiments described in this disclosure may be applicable to any type of vehicle, such as air vehicles, water vehicles, spacecraft, etc. Other exemplary use cases of some embodiments include industrial robotics, computer vision systems, and autonomous driving.

[0108] 9 is a block diagram illustrating a system for implementing an FMCW lidar system, according to some example embodiments. Controller 911 includes processor 940, computer-readable memory 912, storage 958, and user interface 949 having optional display 952 and keyboard 951, these elements connected via bus 956. For example, user interface 964, which communicates between processor 940 and computer-readable memory 912, receives user input from the surface of user interface 957 or keyboard 953, and acquires and stores image data in computer-readable memory 912.

[0109] The computer 911 can include a power supply 954 depending on the application, or the power supply 954 can be located external to the computer 911 if necessary. The user input interface 957 can be configured to connect to a display device 948 via a bus 956. The display device 948 can include, among other things, a computer monitor, a camera, a television, a projector, or a mobile device. The network interface controller (NIC) 934 can be configured to connect to a network 936 via the bus 956, and image data or other data can be rendered on, among other things, a third-party display device, a third-party imaging device, and / or a third-party printing device external to the computer 911.

[0110] 9 , image data or other electronic data may be transmitted via a communication channel of the network 936 and stored in a storage system 958 for storage and / or further processing, among other things. Time-series data or other data may also be received wirelessly or via a wired connection from a receiver 946 (or an external receiver 938) and transmitted wirelessly or via a wired connection via a transmitter 947 (or an external transmitter 939). Both the receiver 946 and the transmitter 947 are connected via a bus 956. The computer 911 may be connected to an external sensing device 944 and an external input / output device 941 via an input interface 908. For example, the external sensing device 904 may include a sensor that collects data before, during, or after collecting the machine's time-series data. The computer 911 may also be connected to another external computer 942. An output interface 909 can be used to output data processed by the processor 940. Additionally, a user interface 949 in communication with the processor 940 and the non-transitory computer-readable storage medium 912 obtains and stores area data in the computer-readable storage medium 912 upon receiving user input from the surface of the user interface 949 .

[0111] The above description provides exemplary embodiments only and is not intended to limit the scope, application, or configuration of the present disclosure. Rather, the following description of exemplary embodiments will provide those skilled in the art with an enabling description for implementing one or more exemplary embodiments. Various changes may be made in the function and arrangement of elements without departing from the spirit and scope of the subject matter as set forth in the appended claims.

[0112] In the following description, specific details are provided to provide a thorough understanding of the embodiments. However, those skilled in the art will understand that embodiments may be practiced without these specific details. For example, systems, processes, and other elements of the disclosed subject matter may be shown as components in block diagrams so as not to obscure the embodiments in unnecessary detail. Also, well-known processes, structures, and techniques may be shown without unnecessary detail so as not to obscure the embodiments. Furthermore, like reference numbers and names in the various drawings refer to like elements. Also, each embodiment may be described as a process that is depicted as a flowchart, flow diagram, data flow diagram, structure diagram, or block diagram. Although a flowchart may describe operations as a sequential process, many operations may be performed in parallel or simultaneously. Also, the order of operations may be changed. A process may be terminated when an operation of a process is completed, although the process may include additional steps not discussed or shown. Moreover, not all operations within a specifically described process need be included in all embodiments. A process may be a method, function, procedure, subroutine, subprogram, etc. If the process is a function, the termination of the function corresponds to the function returning to the calling or main function.

[0113] Furthermore, embodiments of the disclosed subject matter may be implemented, at least in part, manually or automatically. The manual or automated implementation may be implemented, or at least assisted, by machine, hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When implemented in software, firmware, middleware, or microcode, program code or code segments for performing the necessary tasks may be stored on a machine-readable medium. A processor may perform the necessary tasks. The various methods or steps outlined herein may be coded as software executable on one or more processors employing any one of a variety of operating systems or platforms. Furthermore, such software may be written using any of a number of suitable programming languages ​​and / or programming or scripting tools, and compiled as executable machine-language code or intermediate code that runs on a framework or virtual machine. Typically, the functionality of the program modules may be combined or distributed in various embodiments as desired.

[0114] The embodiments of the present disclosure may be embodied as methods, provided by way of example. Operations performed as part of the method may be ordered in any suitable manner. Thus, embodiments may be constructed that perform operations in a different order than the sequential operations in the exemplary embodiments, and that may include performing some operations simultaneously. Furthermore, the use of order terms in the claims, such as first, second, etc., to modify claim elements does not imply a priority, precedence, or order of one claim element relative to another, or a chronological order in which the operations of the method are performed, but is merely used as a label to distinguish the claim elements and (by using order terms) distinguish one claim element with a particular name from another element with the same name. While the present disclosure has been described with reference to certain preferred embodiments, it should be understood that various other modifications and variations can be made within the spirit and scope of the present disclosure. Accordingly, the appended claims are intended to cover all such variations and modifications that fall within the true spirit and scope of the present disclosure.

Claims

1. 1. A frequency modulated continuous wave (FMCW) device comprising: an emitter configured to transmit at least one radiation wave into a scene, the transmitted wave being modulated in the frequency domain using a linear modulation that is subject to impairments that cause nonlinearity in the transmitted wave in the frequency domain; a receiver configured to receive reflections of the transmitted waves from the scene; a mixer operatively connected to the emitter and the receiver and configured to interfere a copy of the transmitted wave with the received reflected wave of the transmitted wave to generate a beat signal; an analog-to-digital converter (ADC) operatively connected to the mixer and configured to generate a series of samples of the beat signal to have a wrapped phase in the time domain; and a processor configured to iteratively estimate the frequency of the beat signal in the time domain based on: 1) phase unwrapping the samples of the beat signal subject to correlated phase errors derived from phase noise statistics of the emitter, until a termination condition is met; and 2) a linear regression fitting the frequency of the beat signal to the unwrapped phase of the beat signal.

2. To perform the current iteration of the estimation, the processor: is configured to determine, for each sample of the series of samples of the beat signal, a current phase error and a phase unwrapping number fitting to a previous frequency of the beat signal and a previous phase offset of the beat signal determined during a previous iteration, wherein the current phase error of the current sample of the series of samples of the beat signal is correlated with a previous phase error of a previous sample of the series of samples of the beat signal by a predetermined phase noise statistic; and 2. The FMCW apparatus of claim 1, configured to update a current frequency of the beat signal and a current phase offset of the beat signal during the current iteration based on the determined current phase error and the determined phase unwrapping number.

3. 3. The FMCW apparatus of claim 2, wherein the processor is configured to iteratively estimate the frequency of the beat signal in the time domain using an alternating optimization including a Viterbi algorithm that probabilistically determines the current phase error and the phase unwrapping number to maximize a fitting likelihood over the series of samples of the beat signal.

4. The FMCW apparatus of claim 3 , wherein the alternating optimization uses generalized least squares (GLS) regression to update the frequency of the beat signal and the phase offset of the beat signal.

5. The FMCW apparatus of claim 3 , wherein the alternating optimization uses least squares regression to update the frequency of the beat signal and the phase offset of the beat signal.

6. 4. The FMCW apparatus of claim 3, wherein the termination condition compares the likelihood provided by the Viterbi algorithm to a threshold.

7. 4. The FMCW apparatus of claim 3, wherein the Viterbi algorithm uses the determined previous phase error and the previous phase unwrapping number of the previous sample to perform a causal estimation of a current phase error and a current phase unwrapping number of each current sample of the series of samples of the beat signal.

8. The FMCW apparatus of claim 7 , wherein the causal estimates of the current phase error and the current phase unwrapping number are determined via linear minimum mean square error estimation.

9. 2. The FMCW apparatus of claim 1, wherein the phase noise statistics include an autocorrelation function of the emitter phase noise and an autocorrelation function of the receiver additive noise.

10. The FMCW device of claim 1 , wherein the termination condition is an iteration count.

11. A lidar comprising the FMCW device of claim 1.

12. The FMCW apparatus of claim 1 , wherein the processor is further configured to estimate a distance to an object in the scene based on the estimated frequency of the beat signal.

13. Being a rider, 13. The FMCW device of claim 12, A lidar, wherein the emitter includes a laser source having a coherence length shorter than the distance to the object.

14. 1. A method for frequency estimation utilizing a frequency modulated continuous wave (FMCW) device, comprising: transmitting at least one radiation wave into a scene, the transmitted wave being modulated in the frequency domain using a linear modulation that is subject to impairments that cause nonlinearity of the transmitted wave in the frequency domain, the method comprising: receiving a reflection of the transmitted wave from the scene; interfering a copy of the transmitted wave with the received reflected wave of the transmitted wave to generate a beat signal; generating a series of samples of the beat signal to have a wrapped phase in the time domain; 2) iteratively estimating the frequency of the beat signal in the time domain based on: 1) phase unwrapping the samples of the beat signal subject to correlated phase errors resulting from phase noise statistics of the emitter; and 2) linear regression fitting the frequency of the beat signal to the unwrapped phase of the beat signal, until a termination condition is met.

15. The current iteration of the frequency estimation determining, for each sample of the series of samples of the beat signal, a current phase error and a phase unwrapping number fitting to a previous frequency of the beat signal and a previous phase offset of the beat signal determined during a previous iteration, wherein the current phase error of a current sample in the series of the beat signal is correlated with a previous phase error of a previous sample in the series of the beat signal by a predetermined phase noise statistic, and the current iteration of the frequency estimation includes:

15. The frequency estimation method of claim 14, comprising updating a current frequency of the beat signal and a current phase offset of the beat signal during the current iteration based on the determined current phase error and the determined phase unwrapping number.

16. 16. The frequency estimation method of claim 15, wherein the iterative estimation of the frequency of the beat signal in the time domain is based on alternating optimization including a Viterbi algorithm that probabilistically determines the current phase error and the phase unwrapping number so as to maximize fitting likelihood over the series of samples of the beat signal.

17. 17. The frequency estimation method of claim 16, wherein the Viterbi algorithm uses the determined previous phase error and the previous phase unwrapping number of the previous sample to perform a causal estimation of a current phase error and a current phase unwrapping number of each current sample of the series of samples of the beat signal.

18. 18. The frequency estimation method of claim 17, wherein the alternating optimization uses generalized least squares (GLS) regression to update the frequency of the beat signal and the phase offset of the beat signal.

19. 16. The frequency estimation method of claim 15, wherein the alternating optimization uses least squares regression to update the frequency of the beat signal and the phase offset of the beat signal.

20. 1. A non-transitory computer-readable medium having stored thereon instructions executable by a computer for controlling a frequency modulated continuous wave (FMCW) device to perform a frequency estimation method, the frequency estimation method comprising: The method includes transmitting at least one radiation wave into a scene, the transmitted wave being modulated in the frequency domain using a linear modulation that is impaired to cause nonlinearity of the transmitted wave in the frequency domain, the method comprising: receiving a reflection of the transmitted wave from the scene; interfering a copy of the transmitted wave with the received reflected wave of the transmitted wave to generate a beat signal; generating a series of samples of the beat signal to have a wrapped phase in the time domain; and iteratively estimating the frequency of the beat signal in the time domain based on: 1) phase unwrapping the samples of the beat signal subject to correlated phase error resulting from phase noise statistics of the emitter; and 2) linear regression fitting the frequency of the beat signal to the unwrapped phase of the beat signal, until a termination condition is met.

Images