Gm Global Technolgoy Operations, Inc.

Vyyo Ltd.

Other

Amnon Jonas

Oded Bialer

Tom Tirer

David Shapiro

Menashe Shahar

Raul Asia

Igal Kotzer

Yonatan Manor

Eric Karl Wilson

Noa Garnett

Ofer Givati

Jack Bettan

Yaron Fishler

Jehuda Ben Simon

Oren Mizrahi

A vehicle, and a system and method for operating the vehicle. The system includes a sensor and a processor. The sensor obtains a detection set related to an object at a range from the vehicle. The processor is configured to select a convolution path for the detection set based on a range of the object, wherein the convolution path includes one or more convolutional layers and a number of the one or more convolutional layers is dependent on the range of the object, apply the one or more convolutional layers of the selected convolution path to the detection set to generate a filtered data set, and operate the vehicle with respect to the object using the filtered data set.

Spatially invariant 3D convolutional network over spherical coordinate input

A radar system and method include and employ a plurality of substantially identical transceiver sets establishing respective substantially identical, overlapping virtual antenna arrays. A first sub-array of widely spaced virtual antennas provides high angular resolution but high angular ambiguity. A second sub-array of narrowly spaced virtual antennas provides low angular ambiguity but low angular resolution.

High resolution unambiguous radar

A system for detecting and estimating a property of an object based on radar includes a signal generator configured to generate a code sequence for a plurality of transmitters configured to emit radar signals over a selected time frame, the code sequence including a plurality of codes, each code of the plurality of codes having a different code length, each code repeated in the code sequence according to a repetition frequency, and each transmitter configured to emit a radar signal based on the code sequence. The system also includes a receiver configured to detect return signals from reflections of the emitted radar signal, and a processing device configured to estimate a property of an object based on the detected return signals.

Variable code sequences for radar transmission

A vehicle, system, and method of determining a velocity of an object. The system includes a radar system and a processor. The radar system is configured to obtain a radar signal with respect to the object over an integration interval, the radar signal including a plurality of velocity samples. The processor is configured to partition the integration interval into a plurality of time segments, each time segment including a subset of the velocity samples, perform a first integration of the subset of the velocity samples within a selected time segment using a first set of velocity of hypotheses to obtain a first stage integration value for the time segment, perform a second integration using the first stage integration value using a second set of velocity hypotheses to obtain a second stage integration value over the integration interval, and determine the velocity of the object from the second stage integration value.

Long signal integration of multiple velocity hypotheses

A vehicle includes a plurality of transmitters of a code division multiple access (CDMA) radar system to simultaneously transmit a frame of transmit signals. A first time duration between transmissions of a first pair of sequential ones of the transmit signals is linearly increased to a second time duration between transmissions of a second pair of sequential ones of the transmit signals. The vehicle also includes a receiver of the CDMA radar system to receive reflected energy resulting from reflection of one of more of the transmit signals of one or more of the plurality of transmitters by an object. A controller processes the reflected energy to obtain information about the object and to control an operation of the vehicle based on the information.

Transmission scheme for implementing code division multiple access in a radar system

A vehicle, radar system for a vehicle and a method of detecting a parameter of an object is disclosed. The radar system includes a base radar and a frequency converter. The base radar generates a first frequency source signal within a first frequency range and is receptive to a first frequency reflected signal within the first frequency range. The base radar is configured to determine a parameter of an object from the first frequency reflected signal. The frequency converter is configured to convert the first frequency source signal to a second frequency source signal within a second frequency range and to convert a second frequency reflected signal within the second frequency range to the first frequency reflected signal.

Generating a high frequency FMCW radar from low frequency FMCW radar

A method of designing a radar system includes implementing a supervised learning process of a neural network to determine a weight corresponding with each of a plurality of patch antennas of a radar system. Each of the plurality of patch antennas is sized in accordance with the weight.

Neural network-based radio frequency network design in a radar system of a vehicle

A method of implementing frequency division multiple access (FDMA) in a radar system of a vehicle includes transmitting a chirp signal from each of a plurality of transmit elements of the radar system simultaneously. The chirp signal transmitted by each of the plurality of transmit elements increases or decreases linearly in frequency over a frequency range over a duration of time and the frequency range of the chirp signal transmitted by adjacent ones of the plurality of transmit elements partially overlapping. The method also includes processing a reflection received based on reflection of the chirp signal transmitted by the plurality of transmit elements by one or more objects and controlling an operation of the vehicle based on locating the one or more objects.

Frequency division multiple access in vehicle radar system

Vehicles and methods for tracking an object and controlling a vehicle based on the tracked object. A Radar-Doppler (RD) map is received from the radar sensing system of the vehicle and relative acceleration of an object with respect to the vehicle is detected based on the RD map so as to provide acceleration data. A current frame of detected object data is received from a sensing system of the vehicle. When the relative acceleration has been detected, a tracking algorithm is adapted to reduce the influence of the predictive motion model or the historical state of the object and the object is tracked using the adapted tracking algorithm so as to provide adapted estimated object data based on the object tracking. One or more vehicle actuators are controlled based on the adapted estimated object data.

Target tracking during acceleration events

A system for estimating a parameter of an object includes a receiver configured to detect a return signal of a radar signal, and a processing device configured to sample the return signal to generate a series of signal samples, partition a time frame into a plurality of successive segments k, and for each segment k, apply a Doppler Fourier transform and calculate a complex value yas a function of Doppler frequencies f. The processing device is also configured to calculate an index based on an acceleration hypothesis and a velocity hypothesis of a set of hypotheses, and for each segment, select one or more Doppler frequency bins based on the index and extract components of the complex value y(f) associated with each selected Doppler frequency bin. The processing device is further configured to calculate a velocity and acceleration spectrum, and estimate an object parameter based on the spectrum.

Radar detection and parameter estimation of accelerating objects

A radar system and method include a sparse array receive element and a processing device. The system performs a beamforming operation on a received radar signal to generate a beamforming spectrum, which contains superposed impulse responses with relative power and angle. The processing device executes an iterative detection routine, starting with a first stage detection that compares the beamforming spectrum to an active power threshold and identifies tentative detection points. In the second stage detection, the processing device determines a certain detection point with the greatest relative power and updates the active power threshold for subsequent iterations of the detection routine. The update involves centering the impulse response related to the certain detection point around its angle, multiplying the relative power by the impulse response, and summing the product with the active power threshold. This process continues until a final set of detection points is obtained.

Robust reflection point detection

Systems and methods to perform object surface estimation using a radar system involve receiving reflected signals resulting from reflection of transmit signals by an object. The method includes processing the reflected signals to obtain an image. The image indicates an intensity associated with at least one set of angle values and a set of range values. The method also includes processing the image to provide the object surface estimation. The object surface estimation indicates a subset of the at least one set of angle values and associated ranges within the set of range values.

Neural network-based object surface estimation in radar system

A system for estimating a range and a velocity of an object includes a processing device configured to perform, for each return pulse of a return signal including reflections of a radar signal, applying a first Fourier transform to the return pulse to transform the return pulse into a range spectrum and calculate a range intensity value for each of a plurality of range hypotheses, calculating a range variation for each of a plurality of hypothesized Doppler frequency values, and for each hypothesized Doppler frequency value, applying a second Fourier transform to the series of return pulses based on the range intensity values and the range variation. The processing device is further configured to perform outputting range and Doppler frequency data including a range-Doppler intensity value for each range hypothesis and hypothesized Doppler frequency, and estimating a range and a velocity of the object based on the range-Doppler intensity values.

Object position and movement estimation using radar

A method and system involve obtaining reflected signals in a radar system using a first one-dimensional array of antenna elements and a second one-dimensional array of antenna elements. The reflected signals result from reflection of transmitted signals from the radar system by one or more objects. The method includes processing the reflected signals obtained using the first one-dimensional array of antenna elements to obtain a first array of angle of arrival likelihood values in a first plane, and processing the reflected signals obtained using the second one-dimensional array of antenna elements to obtain a second array of angle of arrival likelihood values. A four-dimensional image indicating a range, relative range rate, the first angle of arrival, and the second angle of arrival for each of the one or more objects is obtained.

Azimuth and elevation radar imaging with single-dimension antenna arrays of radar system

Systems and methods include transmitting transmit signals from transmit elements, and receiving reflections resulting from the transmit signals at receive elements. The reflections are processed to obtain range-Doppler maps. Each range-Doppler map corresponds with one combination of the transmit elements and the receive elements. The range-Doppler map includes complex values that indicate intensity over a set of range values and a set of relative velocity values. A synthetic matrix of synthetic vectors of array response combinations is generated for transmit angles and receive angles. Each array response combination is a combination of a transmit response for one of the transmit angles and a receive response for one of the receive angles. Two stages of detection are performed. A first stage identifies potential objects and a second stage eliminates the potential objects that are ghost objects. The potential objects remaining after the second stage are the real objects.

Growing community of inventors

Jerusalem, Israel

Amnon Jonas